ISO 22753:2021
(Main)Molecular biomarker analysis - Method for the statistical evaluation of analytical results obtained in testing sub-sampled groups of genetically modified seeds and grains - General requirements
Molecular biomarker analysis - Method for the statistical evaluation of analytical results obtained in testing sub-sampled groups of genetically modified seeds and grains - General requirements
This document describes general requirements, procedures and performance criteria for evaluating the content of genetically modified (GM) seeds/grains in a lot by a group testing strategy that includes qualitative analysis of sub-sampled groups followed by statistical evaluation of the results. This document is applicable to group testing strategy estimating the GM content on a percentage seed/grain basis for purity estimation, testing towards a given reject/accept criterion and for cases where seed/grain lots are carrying stacked events. This document is not applicable to processed products. NOTE Description of the use of group testing strategy are available in References [1], [7], [8], [18], [19] and [20].
Analyse moléculaire de biomarqueurs — Méthode pour l'évaluation statistique des résultats d'analyse obtenus lors des essais de sous-échantillons multiples de semences et de graines génétiquement modifiées — Exigences générales
Le présent document décrit les exigences générales, les modes opératoires et les critères de performance applicables à l’évaluation de la teneur en semences/graines génétiquement modifiées (GM) dans un lot par une stratégie d’analyse de groupe qui comprend l’analyse qualitative de sous-échantillons multiples puis l’évaluation statistique des résultats. Le présent document est applicable à la stratégie d’analyse de groupe permettant d’estimer la teneur en OGM sur un pourcentage de semences/graines afin d’en estimer la pureté, d'évaluer si un critère de rejet/d'acceptation défini est respecté et de déterminer les cas où des lots de semences/graines contiennent un empilement d’événements. Le présent document n’est pas applicable aux produits transformés. NOTE Une description de l’utilisation de la stratégie d’analyse de groupe est donnée dans les Références [1], [7], [8], [18], [19] et [20].
General Information
Overview
ISO 22753:2021 - "Molecular biomarker analysis - Method for the statistical evaluation of analytical results obtained in testing sub‑sampled groups of genetically modified seeds and grains - General requirements" - defines a group testing strategy and statistical framework for estimating the proportion of genetically modified (GM) seeds/grains in a lot. The standard covers preparation of seed/grain groups, qualitative detection of molecular biomarkers (e.g., DNA targets), and statistical evaluation of positive/negative group results to estimate GM content on a percentage seed/grain basis. It is explicitly intended for whole seeds/grains (not processed products) and is suitable for purity estimation, accept/reject decisions and lots containing stacked events.
Key topics and technical requirements
- Group testing principle: divide a representative test sample into predetermined group sizes; perform qualitative tests on each group and use the count of positive groups for statistical inference.
- Sampling and representativeness: requirements for preparing laboratory and test samples so they represent the seed/grain lot.
- Design of testing plans: single‑stage and double‑stage testing plan concepts, specifying group size, number of groups and reject/accept criteria.
- Performance criteria: treatment of false positive rate (FPR), false negative rate (FNR), limit of detection (LOD) and operating characteristic (OC) curves for producer (alpha) and consumer (beta) risk balancing.
- Detection methods: selection and validation of qualitative molecular biomarker assays (e.g., PCR‑based assays) appropriate for group analysis. The standard notes that quantitative PCR of bulk material can overestimate seed counts in the presence of stacked events, motivating the group testing approach.
- Interpretation and reporting: classification of lots as “accept” or “reject”, estimation of GM level in seed/grain lots, and required contents of the test report.
- Informative annexes: worked examples, LOD estimation for testing plans, and experimental methods for determining maximum group size.
Practical applications
- Estimating GM seed/grain purity for seed producers and distributors.
- Conformity testing against regulatory or contractual thresholds and quality limits (AQL/LQL).
- Pre‑shipment and incoming inspection testing in agribusiness supply chains.
- Situations where stacked events may distort quantitative bulk measurements and where whole‑seed counting is required.
Who should use this standard
- Accredited testing laboratories performing molecular biomarker analysis on seeds and grains.
- Seed producers, grain traders and quality assurance teams needing statistically defensible GM content estimates.
- Regulatory authorities and conformity assessment bodies overseeing labeling and threshold compliance.
Related standards
- ISO 16577 - Terms and definitions (molecular biomarker analysis)
- ISO 21572 - Immunochemical methods for detection/quantification of proteins
- ISO 24276 - General requirements for detection of GMOs and derived products
Keywords: ISO 22753:2021, group testing, genetically modified seeds, GM content, statistical evaluation, molecular biomarker analysis, PCR, limit of detection, testing plan, stacked events.
Frequently Asked Questions
ISO 22753:2021 is a standard published by the International Organization for Standardization (ISO). Its full title is "Molecular biomarker analysis - Method for the statistical evaluation of analytical results obtained in testing sub-sampled groups of genetically modified seeds and grains - General requirements". This standard covers: This document describes general requirements, procedures and performance criteria for evaluating the content of genetically modified (GM) seeds/grains in a lot by a group testing strategy that includes qualitative analysis of sub-sampled groups followed by statistical evaluation of the results. This document is applicable to group testing strategy estimating the GM content on a percentage seed/grain basis for purity estimation, testing towards a given reject/accept criterion and for cases where seed/grain lots are carrying stacked events. This document is not applicable to processed products. NOTE Description of the use of group testing strategy are available in References [1], [7], [8], [18], [19] and [20].
This document describes general requirements, procedures and performance criteria for evaluating the content of genetically modified (GM) seeds/grains in a lot by a group testing strategy that includes qualitative analysis of sub-sampled groups followed by statistical evaluation of the results. This document is applicable to group testing strategy estimating the GM content on a percentage seed/grain basis for purity estimation, testing towards a given reject/accept criterion and for cases where seed/grain lots are carrying stacked events. This document is not applicable to processed products. NOTE Description of the use of group testing strategy are available in References [1], [7], [8], [18], [19] and [20].
ISO 22753:2021 is classified under the following ICS (International Classification for Standards) categories: 67.050 - General methods of tests and analysis for food products. The ICS classification helps identify the subject area and facilitates finding related standards.
You can purchase ISO 22753:2021 directly from iTeh Standards. The document is available in PDF format and is delivered instantly after payment. Add the standard to your cart and complete the secure checkout process. iTeh Standards is an authorized distributor of ISO standards.
Standards Content (Sample)
INTERNATIONAL ISO
STANDARD 22753
First edition
2021-08
Corrected version
2022-11
Molecular biomarker analysis —
Method for the statistical evaluation of
analytical results obtained in testing
sub-sampled groups of genetically
modified seeds and grains — General
requirements
Analyse moléculaire de biomarqueurs — Méthode pour l'évaluation
statistique des résultats d'analyse obtenus lors des essais de sous-
échantillons multiples de semences et de graines génétiquement
modifiées — Exigences générales
Reference number
© ISO 2021
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on
the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below
or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii
Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Principle . 4
4.1 General . 4
4.2 Preparation of seed/grain groups . 4
4.3 Detection methods for the qualitative analysis of GM seed/grain in seed/grain
groups . 5
4.4 Statistical evaluation . 5
5 Reagents . 6
6 Apparatus and equipment . 6
7 Design of testing plan . 6
7.1 General . 6
7.2 Single-stage testing plan . 6
7.3 Double-stage testing plan . 7
8 Selection of qualitative methods . 8
8.1 General . 8
8.2 Performance criteria . 8
9 Interpretation .8
10 Expression of results .10
10.1 Classification of a seed/grain lot into “accept” or “reject” category . 10
10.2 Estimation of the level of molecular biomarker in the seed/grain lot . 10
11 Test report .10
Annex A (informative) Terms and definitions comparison table .12
Annex B (informative) Implementation of the method to evaluate GMO content in seeds/
grains example .14
Annex C (informative) Estimation of the limit of detection for a testing plan to detect GM
seeds/grains in seed lots .21
Annex D (informative) Experimental determination of maximum group size .24
Bibliography .25
iii
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www.iso.org/directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www.iso.org/patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to
the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see
www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 34, Food products, Subcommittee SC 16,
Horizontal methods for molecular biomarker analysis.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.
This corrected version of ISO 22753:2021 incorporates the following corrections:
— Formula C.1 has been corrected.
iv
Introduction
Seed and grain testing is used throughout the world to commercially define the purity of seed and grain
lots.
Commercial requirements for labelling agricultural products with genetically modified organism (GMO)
content at a specified threshold level both as a seed/grain contaminant and a food ingredient have
become common to satisfy regulations and consumer demands. Conformance with these specifications
is evaluated at various points of the supply chain, often starting with the harvested grain.
Quantitative real-time polymerase chain reaction (PCR) can be used to determine the GMO content by
analysis of the ratio of GMO DNA copy numbers to plant-species specific DNA copy numbers followed by
a conversion to genetically modified (GM) mass fraction.
Multiple events stacked in a crop, such as those generated by crossing two or more single events,
are widely used in agricultural production. A stacked event seed or grain containing GMO DNA
corresponding to two or more GM events commingled in lot cannot be differentiated by quantitative
PCR alone from multiple seeds within the lot each containing a single GM event. Consequently, if the
actual measured GMO arises only from GM stacked event seeds, GM content measured by quantitative
real-time PCR of a single sample will lead to an overestimation of the actual number of GM seeds or
grains present.
The group testing strategy described in this document provides a reliable alternative to estimate the
GM content on the basis of the fact that whole seeds/grains are the sample material.
The process described in this document can provide a method to accurately estimate the percentages
of GM seeds/grains in a lot irrespective of the presence of stacked event seeds/grains. GM content is
determined for representative subsampled groups of seed/grain from a lot and statistically analysed.
v
INTERNATIONAL STANDARD ISO 22753:2021(E)
Molecular biomarker analysis — Method for the statistical
evaluation of analytical results obtained in testing sub-
sampled groups of genetically modified seeds and grains —
General requirements
1 Scope
This document describes general requirements, procedures and performance criteria for evaluating
the content of genetically modified (GM) seeds/grains in a lot by a group testing strategy that includes
qualitative analysis of sub-sampled groups followed by statistical evaluation of the results.
This document is applicable to group testing strategy estimating the GM content on a percentage seed/
grain basis for purity estimation, testing towards a given reject/accept criterion and for cases where
seed/grain lots are carrying stacked events.
This document is not applicable to processed products.
NOTE Description of the use of group testing strategy are available in References [1], [7], [8], [18], [19] and
[20].
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content
constitutes requirements of this document. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any amendments) applies.
ISO 16577, Molecular biomarker analysis — Terms and definitions
ISO 21572, Foodstuffs — Molecular biomarker analysis — Immunochemical methods for the detection and
quantification of proteins
ISO 24276, Foodstuffs — Methods of analysis for the detection of genetically modified organisms and
derived products — General requirements and definitions
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 16577 and the following apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at https:// www .electropedia .org/
3.1
absolute PCR limit of detection
absolute polymerase chain reaction limit of detection
absolute PCR LOD
lowest nominal (average) number of target copies in the template volume distributed to individual PCRs
that would allow for an acceptable probability of detecting the target
3.2
AQL
A
QL
acceptable quality limit
level of impurity that is acceptable to the producer and that production practices can support
3.3
consumer risk
consumer (beta) risk
probability of accepting a lot at the lower quality limit (3.10)
3.4
deviant seed/grain
considered non-conforming based on the presence or absence of a specific trait or characteristic
Note 1 to entry: For the purpose of this document, a deviant seed is considered to possess a GM characteristic
that is not expected or is unintended based on the expected or known GM characteristics of the seed/grain.
3.5
false negative rate
FNR
probability that a known positive (seed/grain group) test sample (3.20) has been classified as negative
by the method
Note 1 to entry: The false negative rate is the number of misclassified known positives divided by the total
number of positive test samples (3.20).
[SOURCE: ISO 16577:2016, 3.63, modified — the abbreviation has been added, “positive test sample”
has been changed to “positive (seed/grain group) test sample”, and the formula has been deleted.]
3.6
false positive rate
FPR
probability that a known negative (seed/grain group) test sample (3.20) has been classified as positive
by the method
Note 1 to entry: The false positive rate is the number of misclassified known negatives divided by the total
number of negative test samples (3.20).
[SOURCE: ISO 16577:2016, 3.65, modified — the abbreviation has been added, “negative test sample”
has been changed to “negative (seed/grain group) test sample”, and the formula has been deleted.]
3.7
group size
number of seeds/grains comprising a group
3.8
group testing
statistical evaluation of analyte contents based on qualitative analysis results (i.e. positive or negative)
from each seed/grain group in the test sample (3.20)
3.9
laboratory sample
sample or subsample(s) received by the laboratory
Note 1 to entry: The seed/grain sample received is expected to represent the seed/grain lot (3.18).
[SOURCE: ISO 16577:2016, 3.89, modified — Note 1 to entry has been added.]
3.10
LQL
L
QL
lower quality limit
highest impurity that is acceptable to the consumer
Note 1 to entry: This can be equivalent to the threshold (3.22).
3.11
mass fraction
ratio of GM seeds/grains relative to the total seeds/grains corresponding to mass ratio
3.12
number of deviant seed/grain groups
number of seed/grain groups (3.17) including one or more deviant seeds/grains (3.4)
3.13
operating characteristic curve
OC curve
graph plotting the percentage of deviant seeds/grains and the probability of acceptance respectively
on the horizontal and the vertical axes and used in quality control to determine the probability of
accepting seed/grain lots (3.18) in a testing plan (3.21)
3.14
producer risk
producer (alpha) risk
probability of rejecting a lot at the AQL (3.2)
3.15
representative sample
sampling units (samples or groups) that have been extracted from a lot with the process ensuring all
sampling units of the lots have an equal probability of being selected and not altered in any way that
would change the analytical result
Note 1 to entry: The extraction process can be a multi-stage process.
3.16
reject/accept criterion
maximum number of deviant seed/grain groups (3.12) that can be detected in the test sample (3.20) of an
acceptable seed/grain lot (3.18)
3.17
seed/grain group
group
determined number of seeds/grains prepared from a seed/grain test sample (3.20) by representative
sampling
3.18
seed/grain lot
lot
population for which sampling is intended to estimate the measured parameter
3.19
stacked event
accumulation of two or more transformation events as a result of traditional breeding and/or successive
transformation steps)
Note 1 to entry: In the context of this document a stacked event refers to a stack in which the two or more events
are not genetically linked.
[SOURCE: ISO 16577:2016, 3.197, modified — Note 1 to entry has been added.]
3.20
test sample
sample prepared for testing or analysis, the whole quantity or part of it being used for testing or
analysis at one time
Note 1 to entry: The test sample is prepared from the laboratory sample (3.9).
Note 2 to entry: The test sample is expected to represent the laboratory sample (3.9).
[SOURCE: ISO 16577:2016, 3.210, modified — Note 1 to entry and Note 2 to entry have been added.]
3.21
testing plan
plan specifying group testing (3.8) conditions including group size (3.7), the number of seed/grain groups
(3.17) and the number of deviant seed/grain groups (3.12) in test sample (3.20) resulting in rejection of
seed/grain lot (3.18)
3.22
threshold
maximum acceptable content of GMO presence in a lot
Note 1 to entry: This can be a prescribed value.
Note 2 to entry: Thresholds can be expressed in mass fraction (3.11) with the proviso that an uncertainty factor
is involved in the conversion to a seed/grain percentage threshold.
4 Principle
4.1 General
In this method, the test sample is divided into a predetermined number of groups. Each group consists
of a determined number of seed/grain and is tested qualitatively for the presence or absence of a GM
target. A statistical evaluation is performed on the number of GM positive groups relative to the total
number of seed/grain groups to determine the GM content in mass fraction.
A statistical calculation determines the optimal testing conditions, namely, the number of seeds/grains
per group (group size), the number of seed/grain groups, and the maximum number of GMO positive
seed/grain groups for seed/grain lot acceptance. Alternatively, a statistical calculation provides an
estimate of the percentage by number of the GM seeds/grains in a lot, according to a given testing plan.
4.2 Preparation of seed/grain groups
Key
1 bulk seed/grain lot
2 laboratory sample
3 test sample
4 seed/grain groups
5 deviant seed/grain
NOTE Each group is represented as an array on the right.
Figure 1 — Sampling illustration of the obtention of seed/grain groups from a bulk seed/grain
lot
The process of forming seed/grain groups from a series of sampling steps starting with the bulk seed/
grain lot is shown in Figure 1, (1).
Although the procedures for obtaining a laboratory sample from a seed/grain lot is not the subject
of this document, a laboratory sample (2) from a seed/grain lot shall be obtained appropriately. The
procedures can be designed according to the References [3], [6], [10], [11], [12], [15], [19] and [23].
The laboratory sample shall be thoroughly mixed and divided/reduced to create the test sample (3).
Likewise, the test sample shall be thoroughly mixed (i.e. homogeneous) and divided into seed/grain
groups (each group represented as an array in Figure 1, (4)) following simple random sampling
principles. The seed/grain groups can vary in size from one single seed/grain up to the complete test
sample (i.e. a single bulk). In most cases, multiple seed/grain groups are created from the test sample.
A determined number of seeds/grains can either be obtained by weighing or a volumetric measurement,
where an approximation of number is made based on a determined conversion factor (e.g. thousand
seeds/grains weight). For the case that weight is used to obtain the seed/grain groups, the operator
shall have an estimate of the variability introduced by using weight rather than seed/grain count.
The group testing procedure described in Clause 7 is carried out on the collective qualitative (positive
or negative) results for each seed/grain group.
4.3 Detection methods for the qualitative analysis of GM seed/grain in seed/grain
groups
[21]
In general, GMO detection methods are categorized into two classes . The first class of assays targets
a nucleic acid sequence for detecting GMO presence. The second class includes methods for detecting a
specified protein that confers a specific transgenic trait. Detection methods from either or both classes
should be selected considering fitness-for-purpose. Guidance on the selection of qualitative methods is
[4]
provided in Clause 8. Further details can be found in ISO 21569 and ISO 21572.
4.4 Statistical evaluation
Sampling and measurement uncertainty shall be considered. Sampling uncertainty can be adequately
[18][2]
considered using the binomial distribution . The FPR and the FNR of the qualitative assay should
[2]
be considered . The LOD of the applied detection method should be considered.
The group testing described here can be used to set reject/accept criteria based on a given threshold by
GMO content, as well as to estimate the GMO content and associated upper and lower confidence limits.
5 Reagents
All reagents used in the analysis should be those specified in the method.
Otherwise, all reagents should be of molecular biology grade.
These reagents shall be stored and used as recommended by the supplier or according to the laboratory
quality assurance specifications. It can also be appropriate to aliquot the reaction solutions required
for the analytical method in order to avoid subjecting them to repeated freeze–thaw cycles, or to reduce
the chances of cross contamination or both. Further details shall refer to ISO 24276 and ISO 21572.
6 Apparatus and equipment
The laboratory should use properly maintained equipment suitable for the methods employed.
Further details shall refer to ISO 24276 and ISO 21572.
7 Design of testing plan
7.1 General
The number of seeds/grains tested, the reject/accept criteria, the sample preparation steps and the
method used for testing shall be determined depending on the analytical purpose.
In seed/grain sample classification, it can be determined whether the number of deviant seeds/grains
or seed/grain groups is above a given reject/accept criterion or not. Then, it can be decided to reject or
accept the seed/grain lot based on the test results.
A basic testing plan for group testing consists of three fundamental parameters:
a) the number of seed/grain groups;
b) the size of the seed/grain groups;
c) the maximum number of deviant seed/grain groups for seed/grain lot acceptance (reject/accept
criterion).
The risks associated with the AQL and the LQL are the producer (alpha) and consumer (beta) risks
respectively, and together with the FPR and FNR allow the design of an appropriate testing plan.
The OC curve can be used to develop a testing plan. Explanations for the estimation of the LOD for a
zero deviant testing plan, the effect of the genome size on the group size if methods targeting DNA are
applied, and the effect of the individual seed size on the sample preparation are given in Annex C.
Annex D provides guidance on the determination of the maximum group size whatever analytical
method is used in the laboratory.
[16]
NOTE Seedcalc is a statistical program (Microsoft Excel spreadsheet application) that is freely available
from the International Seed Testing Association and has procedures to facilitate the design. Seedcalc is located
on the ISTA website.
7.2 Single-stage testing plan
A single-stage testing plan consists of one testing stage. Groups are taken from the test sample and
evaluated once, and a decision is then made based on the results to accept or reject the seed/grain
test sample. In a single-stage testing plan, a specified number of individual seeds/grains or seed/
grain groups shall be selected randomly from the test sample and tested. Depending on the number
of deviants detected and the maximum number of deviants specified in the plan, the seed/grain lot is
either accepted or rejected.
The probability that an individual seed/grain or seed/grain group is deviant, p , can be calculated as
b
given in Formula (1):
m
pP=−11=−()1−p (1)
b
where
P is the probability that there are no deviant seeds/grains in the group;
p is the true unknown impurity in the seed/grain lot;
m is the number of individual seeds/grains in a seed/grain group (if seeds/grains are tested indi-
vidually, m = 1).
Then, the probability that a lot will be accepted, P(a) is calculated as given in Formula (2):
c
n
ni−
i
P()a = pp()1− (2)
∑ b b
i
i=0
where
P(a) is the probability that a lot will be accepted;
n is the number of individual seeds/grains or seed/grain groups tested;
c is the maximum number of deviant seed/grain groups for acceptance.
By combining Formulae (1) and (2), P(a) is a function of p, n, m and c.
After n, m and c are determined, an OC curve can be drawn by plotting p and P(a) on the x-axis and
y-axis, respectively.
7.3 Double-stage testing plan
A double-stage testing plan is generally set up so that additional seed/grain groups are tested in the
second stage. Initial seed/grain groups are taken from the test sample and tested. Based on this test
result, three different decisions can be made:
a) accept the seed/grain lot;
b) reject the seed/grain lot; or
c) draw a second set of seed/grain groups from the test sample and retest.
The test results from the first and second stages of testing are combined and used to determine whether
the seed/grain lot is accepted or rejected (see Figure B.1). In Annex B examples for implementation of a
double-stage testing plan to evaluate GMO content in seeds/grains are provided. Subclause B.1 can also
be applied for cases where seed/grain lots are carrying stacked events.
Some additional terms are defined as follows:
— n , the number of independent seed/grain groups to be tested in the first stage;
— n , the number of independent seed/grain groups to be tested in the second stage;
— c , the maximum number of allowable deviant seed/grain groups for acceptance in the first stage;
— c , the minimum number of deviant seed/grain groups that will result in rejection at the first stage;
— c , the maximum number of deviant seed/grain groups in the first and second stages combined
allowed for acceptance;
— d , the number of deviant seed/grain groups in the first stage;
— d , the number of deviant seed/grain groups in the second stage.
P(a) is calculated as given in Formula (3):
n n n
c c −1 ci-
11ni−−ni 2 nj−
1 i 2 i 3 j
11 2
P()a = pp()11− + pp()− × pp()1−
∑ b b ∑ b b ∑ b b
i=0 ic=+1 j=0
i i j
(3)
8 Selection of qualitative methods
8.1 General
An analytical method shall be chosen to meet the purpose of testing. The performance characteristics
of the method should be determined before application in seed/grain testing.
Analytical methods have been developed to detect specific genes encoding transgenic traits or specific
characteristics expressed by specific genes in seeds/grains. Nucleic-acid-based methods such as PCR
[4][5]
are available that detect specific DNA sequences encoding elements, constructs or GMO events .
Protein-based methods such as ELISA and lateral flow immunoassays require a specific antibody for
detecting a specific GM protein (see ISO 21572).
8.2 Performance criteria
The analytical methods applied for the test plan protocol shall detect at least one GM seed/grain in a
group with high probability of detection. Refer to Reference [2].
In the case of PCR, detection methods shall be chosen to meet the purpose of group testing. General
methods performance criteria are described in ISO 24276. General criteria for the design of the testing
plan which should be considered include
a) physical and genome size of seed/grain species as it affects the number of seed/grain that can be
easily ground per group and the number of genome equivalents that can be analysed in a standard
PCR, respectively,
b) absolute PCR limit of detection of the qualitative method, and
c) false-negative rates associated with the method of detection or identification in addition for both
[8][34]
nucleic acid- and protein-based methods should be considered .
[4]
Detection-method-specific performance criteria can refer to ISO 24276, ISO 21569 and ISO 21572.
The seed/grain testing plans discussed in this document assume that the seeds/grains tested are a
representative sample drawn from the seed/grain test sample. Simple representative sampling implies
that each seed/grain in the test sample has both an equal and an independent chance of being included
in the seed/grain group.
9 Interpretation
In determining whether to “accept” or “reject” a given seed/grain lot, the test results shall be compared
with the predetermined reject/accept criterion, e.g. the maximum number of GM-positive groups
allowable for acceptance.
Statistical calculation using the formulae shown below permit the evaluation of a GMO content with
1)
confidence intervals from the test results. Statistical calculation programs such as Seedcalc facilitate
the calculation. In this manner, one can obtain quantitative information on the GMO content of the seed/
grain lot based on how many groups proved to be GM-positive in the qualitative analysis. Together, test
results and their statistical evaluation reveal the level of impurity in the seed/grain lot. Ninety-five
percent upper and lower confidence limits for this impurity evaluation can then be calculated. The true
impurity in the seed/grain test sample can be expected with 95 % confidence to fall within these limits.
The most likely value of GMO content, p, can be evaluated from the test results as given in Formula (4).
d
m
p=−11− (4)
n
where
n is the number of individual seeds/grains or seed/grain groups tested;
m is the number of individual seeds/grains in a seed/grain group (if the seeds/grains are tested
individually, m = 1);
d is the number of deviant seeds/grains or seed/grain groups.
The group testing approach, like quantitative methods, has limitations concerning the GM levels that
can be estimated. Table 1 gives two examples of the highest computed GM estimate for test sample sizes
of 200 seeds/grains and 3 000 seeds/grains. These highest estimates are obtained when all, but one
group is positive. Associated 95 % confidence limits are given to the estimates to show the sampling
uncertainty.
Note For seed/grain group sizes greater than one, when all groups are positive for GM presence, there is
very limited utility in this approach.
Table 1 — Examples of highest computed GM estimate of the content of the deviant seeds/grains
for various seed/grain group sizes (when all but one group is positive) and the 95 % confidence
limits (when all but one group is positive)
Seeds per GM positive Estimated percent- Range of GMO content (%)
Seeds (total) Groups
groups groups age GM seed (for 95 % confidence level)
1 200 0 0,0 0,0 to 1,8
5 40 4 3,9 0,8 to 12,4
10 20 9 10,9 4,0 to 25,8
20 10 19 25,9 13,0 to 48,7
1 3 000 0 0,0 0,0 to 0,1
5 600 4 0,3 0,1 to 0,9
10 300 9 0,8 0,3 to 2,0
3 000
20 150 19 2,0 0,9 to 4,4
30 100 29 3,3 1,7 to 6,8
60 50 59 7,9 4,7 to 14,4
1) Seedcalc is an example of a statistical tool for seed testing. This information is given for the convenience of users
of this document and does not constitute an endorsement by ISO of this product.
If the confidence level for evaluation is set at x %, the upper confidence limit of GMO content in the
evaluation can be calculated using the following Formulae (5) and (6):
x
α =−1 (5)
where x is the confidence level in percentage terms.
()dF+1 m
12−+α ,,dn22 −2d
P =−11− (6)
UL
()nd− ++()dF1
12−+α ,,dn22 −2d
where the quantity F is the 1 − α quantile from an F-distribution with 2d + 2 and 2n − 2d
12−+α ,,dn22 −2d
degrees of freedom.
Also, the two-sided confidence interval (upper limit, P ; lower limit, P ) can be calculated using the
UL LL
following Formulae (7) and (8):
dF+1
() m
12−+α/,22dn,22− d
P =−11− (7)
UL
nd− ++dF1
() ()
12−+α/,22dn,22− d
m
d
P =−11− (8)
LL
dn+−()dF+1 /
α/,22dn,22−+d 2
10 Expression of results
10.1 Classification of a seed/grain lot into “accept” or “reject” category
To classify a seed/grain lot into the “accept” or “reject” category, a statement can be made such that the
seed/grain lot is acceptable or that the seed/grain lot should be rejected.
The upper 95 % confidence limit of the concentration based on the result can be included, or the number
of groups tested, or the number of deviant pools or all of these.
The OC curve expressing the characteristic of sampling can be attached along with the alternative
decision result in order to facilitate understanding.
10.2 Estimation of the level of molecular biomarker in the seed/grain lot
The GMO content in the seed/grain lot can be estimated as described in Clause 9. A statement can be
made such that the most probable value of GMO content is p %, and the ()1−×α 100 % confidence
interval ranges from P % to P %.
LL UL
11 Test report
The test report shall be written in accordance with ISO 24276 and shall contain at least the following
additional information:
a) the sample;
b) a reference to the method that was used for the extraction of nucleic acid or protein;
c) a reference to the methods used for the amplification of the nucleic acid target sequences or the
methods used for the detection of the target protein or both;
d) the LOD of the method used to test the groups and the matrix used to identify the LOD;
e) the reference material used if applicable;
f) the results expressed according to Clause 10;
g) the International Standard used (i.e. ISO 22753:2021);
h) any deviations from the procedure;
i) any unusual features observed;
j) the date of the test.
Annex A
(informative)
Terms and definitions comparison table
A.1 Comparison of terms defined other documents
Synonymous terms defined in other documents or organizations are shown in Table A.1.
Table A.1 — Terms comparison table
a,b c
This document ISTA JRC
Term Definition Term Definition Term Definition
a lot is a distinct
and specified
quantity of ma-
population for which a seed lot is a specified
terial dispatched
seed/grain sampling is intended to quantity of seed that is
seed lot lot or received at one
lot estimate the measured physically and uniquely
time and covered
parameter identifiable
by a particular
contract or ship-
ping document
a submitted sample is
a sample that is to be
submitted to the testing
sample or subsample(s)
laboratory and may com- sample as pre-
received by the labo-
prise either the whole of pared (form the
ratory
the composite sample or lot) for sending
laboratory submitted laboratory
Note 1 to entry: The
a subsample thereof. The to the laboratory
sample sample sample
seed/grain sample
submitted sample may be and intended for
received is expected
divided into subsamples inspection or
to represent the seed/
packed in different ma- testing
grain lot.
terial meeting conditions
for specific tests (e.g.
moisture or health).
a
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 2: Sampling. International Rules for Seed Testing 2021, 2021,
[15]
Bassersdorf, Switzerland .
b
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 19: Testing for seeds of genetically modified organisms.
[17]
International Rules for Seed Testing 2021, 2021, Bassersdorf, Switzerland .
c
EUROPEAN COMMISSION, JOINT RESEARCH CENTRE. (2014), JRC Technical Report: Guidelines for sample preparation
[9]
procedures in GMO analysis .
TTaabblle Ae A.1 1 ((ccoonnttiinnueuedd))
a,b c
This document ISTA JRC
Term Definition Term Definition Term Definition
sample prepared for
testing or analysis,
the whole quantity or the working sample is the
part of it being used for whole of the submitted
testing or analysis at sample or a subsam-
(sub-)sample pre-
one time ple thereof, on which
pared from the
one of the quality tests
Note 1 to entry: The working laboratory sample
test sample described in these ISTA test sample
test sample is prepared sample and from which
Rules is made and must
from the laboratory test portions will
be at least the weight
sample. be taken
prescribed by the ISTA
Note 2 to entry: The Rules for the particular
test sample is expected test.
to represent the labora-
tory sample.
A seed group is one of the
portions of the working
sample that is separately
determined number of
prepared (e.g. grinding,
seeds/grains prepared
seed/grain DNA or protein extrac-
from a seed/grain test seed group - -
group tion) and analysed (e.g.
sample by representa-
end-point PCR, ELISA,
tive sampling
real-time PCR) when
using the group testing
approach
a
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 2: Sampling. International Rules for Seed Testing 2021, 2021,
[15]
Bassersdorf, Switzerland .
b
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 19: Testing for seeds of genetically modified organisms.
[17]
International Rules for Seed Testing 2021, 2021, Bassersdorf, Switzerland .
c
EUROPEAN COMMISSION, JOINT RESEARCH CENTRE. (2014), JRC Technical Report: Guidelines for sample preparation
[9]
procedures in GMO analysis .
Annex B
(informative)
Implementation of the method to evaluate GMO content in seeds/
grains example
B.1 Example 1: Group testing to evaluate GMO content in maize grains
B.1.1 General
This annex provides an example to evaluate GMO content in grains using a double-stage testing plan.
The double stage testing plan is used for checking the appropriateness of food labelling in Japan.
B.1.2 Analytical purpose
If there is commingling of stacked event(s) into seed/grain lots, GMO amounts measured by real-time
PCR lead to an overestimation as compared to the actual proportional GM amount in a lot. The group
testing strategy was introduced to estimate the GM content towards the given reject/accept criterion,
irrespective of the presence of stacked event seeds/grains in the lot.
The specified sampling strategy was devised to determine if the GMO content in maize grain lots
exceeds 5 % (mass/mass) or not.
B.1.3 Properties of the analytical method
[19]
Properties for each item are shown in Table B.1. Flow chart for decision making is shown in Figure B.1.
[22][13]
Table B.1 — Analytical method example for double stage sampling plan
Item Value
Type of analytical sample Maize grain
Analyte GMOs in maize grains
Real-time PCR method targeting either or both of Cauliflower
mosaic virus derived 35S promoter and Agrobacterium tumefa-
Qualitative molecular biomarker detection
ciens derived NOS (nopaline synthase) terminator GM elements,
method
which is validated to be able to detect at least 1 GM grain in
20 grains
Type of sampling plan Double stage
Group size in the first stage 20 grains
Number of groups in the first stage 10 groups
6 groups
Reject/accept criterion in the first stage
The second stage is carried out if 7 or more groups show positive.
Group size in the second stage 20 grains
Number of groups in the second stage 10 groups
12 groups after combining the numbers of positive group in the
first and second stages.
Reject/accept criterion in the second stage
Sample lot rejected if ≥ 13 groups test positive for either or both
GM targets.
Figure B.1 — Flow chart for decision making
B.2 Example 2: Group testing to evaluate GMO content in seeds
B.2.1 Analytical purpose
[20]
This example describes a validated German official approach for seed testing . It can be applied to
determine whether the GMO content exceeds 0,1 % or not.
B.2.2 Properties of the analytical method
Properties for each item are shown in Table B.2. Flow chart for decision making is shown in Figure B.2.
The Seedcalc based estimations of GM seed content based on the number of GM positive results obtained
[20][14]
in the qualitative tests are given in Tables B.3 and B.4.
Table B.2 — Analytical method example for German official approach
Item Value
Type of analytical sample Seeds (maize; rapeseed)
Analyte Genetically modified seed
Qualitative molecular biomarker detection Real-time PCR method validated to be able to detect at least 1
method genetically modified seed in 3 000 seeds
Type of sampling plan Double stage
Group size in the first stage 1 000 seeds
Number of groups in the first stage 3 groups
When GMO is detected in more than 1 group: GM seed content
exceeds 0,1 %;
Reject/accept criterion in the first stage
second stage is carried out, if GMO is detected only in 1 group.
Group size in the second stage 1 000 seeds
Number of groups in the second stage 3 groups
In additional group(s) GMO detected: GM seed content exceeds
Reject/accept criterion in the second stage
0,1 %
a
Quantitative real time PCR tests or comparison of Cq values can help to distinguish whole GM seeds present
in the sample from positive signals caused by traces of other homogenously present material (e.g. dust, seed
coating).
Figure B.2 — Flow chart for decision making
Table B.3 — Seedcalc estimations of GM seed content based on the number of GM positive
results obtained in the qualitative tests of three groups of 1 000 seeds
Range of GMO
Seeds GM Percentage GM Probability for a
Seeds
...
INTERNATIONAL ISO
STANDARD 22753
First edition
2021-08
Molecular biomarker analysis —
Method for the statistical evaluation of
analytical results obtained in testing
sub-sampled groups of genetically
modified seeds and grains — General
requirements
Analyse moléculaire de biomarqueurs — Méthode pour l'évaluation
statistique des résultats d'analyse obtenus lors des essais de sous-
échantillons multiples de semences et de graines génétiquement
modifiées — Exigences générales
Reference number
©
ISO 2021
© ISO 2021
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting
on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address
below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii © ISO 2021 – All rights reserved
Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Principle . 4
4.1 General . 4
4.2 Preparation of seed/grain groups . 5
4.3 Detection methods for the qualitative analysis of GM seed/grain in seed/grain groups . 5
4.4 Statistical evaluation . 5
5 Reagents . 6
6 Apparatus and equipment . 6
7 Design of testing plan . 6
7.1 General . 6
7.2 Single-stage testing plan . 6
7.3 Double-stage testing plan . 7
8 Selection of qualitative methods . 8
8.1 General . 8
8.2 Performance criteria . 8
9 Interpretation . 8
10 Expression of results .10
10.1 Classification of a seed/grain lot into “accept” or “reject” category .10
10.2 Estimation of the level of molecular biomarker in the seed/grain lot .10
11 Test report .10
Annex A (informative) Terms and definitions comparison table .12
Annex B (informative) Implementation of the method to evaluate GMO content in seeds/
grains example.14
Annex C (informative) Estimation of the limit of detection for a testing plan to detect GM
seeds/grains in seed lots .21
Annex D (informative) Experimental determination of maximum group size .24
Bibliography .25
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www .iso .org/ directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www .iso .org/ patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to the
World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www .iso .org/
iso/ foreword .html.
This document was prepared by Technical Committee ISO/TC 34, Food products, Subcommittee SC 16,
Horizontal methods for molecular biomarker analysis.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www .iso .org/ members .html.
iv © ISO 2021 – All rights reserved
Introduction
Seed and grain testing is used throughout the world to commercially define the purity of seed and grain
lots.
Commercial requirements for labelling agricultural products with genetically modified organism (GMO)
content at a specified threshold level both as a seed/grain contaminant and a food ingredient have
become common to satisfy regulations and consumer demands. Conformance with these specifications
is evaluated at various points of the supply chain, often starting with the harvested grain.
Quantitative real-time polymerase chain reaction (PCR) can be used to determine the GMO content by
analysis of the ratio of GMO DNA copy numbers to plant-species specific DNA copy numbers followed by
a conversion to genetically modified (GM) mass fraction.
Multiple events stacked in a crop, such as those generated by crossing two or more single events,
are widely used in agricultural production. A stacked event seed or grain containing GMO DNA
corresponding to two or more GM events commingled in lot cannot be differentiated by quantitative
PCR alone from multiple seeds within the lot each containing a single GM event. Consequently, if the
actual measured GMO arises only from GM stacked event seeds, GM content measured by quantitative
real-time PCR of a single sample will lead to an overestimation of the actual number of GM seeds or
grains present.
The group testing strategy described in this document provides a reliable alternative to estimate the
GM content on the basis of the fact that whole seeds/grains are the sample material.
The process described in this document can provide a method to accurately estimate the percentages
of GM seeds/grains in a lot irrespective of the presence of stacked event seeds/grains. GM content is
determined for representative subsampled groups of seed/grain from a lot and statistically analysed.
INTERNATIONAL STANDARD ISO 22753:2021(E)
Molecular biomarker analysis — Method for the statistical
evaluation of analytical results obtained in testing sub-
sampled groups of genetically modified seeds and grains —
General requirements
1 Scope
This document describes general requirements, procedures and performance criteria for evaluating
the content of genetically modified (GM) seeds/grains in a lot by a group testing strategy that includes
qualitative analysis of sub-sampled groups followed by statistical evaluation of the results.
This document is applicable to group testing strategy estimating the GM content on a percentage seed/
grain basis for purity estimation, testing towards a given reject/accept criterion and for cases where
seed/grain lots are carrying stacked events.
This document is not applicable to processed products.
NOTE Description of the use of group testing strategy are available in References [1], [7], [8], [18], [19] and
[20].
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content
constitutes requirements of this document. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any amendments) applies.
ISO 16577, Molecular biomarker analysis — Terms and definitions
ISO 21572, Foodstuffs — Molecular biomarker analysis — Immunochemical methods for the detection and
quantification of proteins
ISO 24276, Foodstuffs — Methods of analysis for the detection of genetically modified organisms and
derived products — General requirements and definitions
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 16577 and the following apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at https:// www .electropedia .org/
3.1
absolute PCR limit of detection
absolute polymerase chain reaction limit of detection
absolute PCR LOD
lowest nominal (average) number of target copies in the template volume distributed to individual PCRs
that would allow for an acceptable probability of detecting the target
3.2
AQL
A
QL
acceptable quality limit
level of impurity that is acceptable to the producer and that production practices can support
3.3
consumer risk
consumer (beta) risk
probability of accepting a lot at the lower quality limit (3.10)
3.4
deviant seed/grain
considered non-conforming based on the presence or absence of a specific trait or characteristic
Note 1 to entry: For the purpose of this document, a deviant seed is considered to possess a GM characteristic
that is not expected or is unintended based on the expected or known GM characteristics of the seed/grain.
3.5
false negative rate
FNR
probability that a known positive (seed/grain group) test sample (3.20) has been classified as negative
by the method
Note 1 to entry: The false negative rate is the number of misclassified known positives divided by the total
number of positive test samples (3.20).
[SOURCE: ISO 16577:2016, 3.63, modified — the abbreviation has been added, “positive test sample”
has been changed to “positive (seed/grain group) test sample”, and the formula has been deleted.]
3.6
false positive rate
FPR
probability that a known negative (seed/grain group) test sample (3.20) has been classified as positive
by the method
Note 1 to entry: The false positive rate is the number of misclassified known negatives divided by the total
number of negative test samples (3.20).
[SOURCE: ISO 16577:2016, 3.65, modified — the abbreviation has been added, “negative test sample”
has been changed to “negative (seed/grain group) test sample”, and the formula has been deleted.]
3.7
group size
number of seeds/grains comprising a group
3.8
group testing
statistical evaluation of analyte contents based on qualitative analysis results (i.e. positive or negative)
from each seed/grain group in the test sample (3.20)
3.9
laboratory sample
sample or subsample(s) received by the laboratory
Note 1 to entry: The seed/grain sample received is expected to represent the seed/grain lot (3.18).
[SOURCE: ISO 16577:2016, 3.89, modified — Note 1 to entry has been added.]
2 © ISO 2021 – All rights reserved
3.10
LQL
L
QL
lower quality limit
highest impurity that is acceptable to the consumer
Note 1 to entry: This can be equivalent to the threshold (3.22).
3.11
mass fraction
ratio of GM seeds/grains relative to the total seeds/grains corresponding to mass ratio
3.12
number of deviant seed/grain groups
number of seed/grain groups (3.17) including one or more deviant seeds/grains (3.4)
3.13
operating characteristic curve
OC curve
graph plotting the percentage of deviant seeds/grains and the probability of acceptance respectively
on the horizontal and the vertical axes and used in quality control to determine the probability of
accepting seed/grain lots (3.18) in a testing plan (3.21)
3.14
producer risk
producer (alpha) risk
probability of rejecting a lot at the AQL (3.2)
3.15
representative sample
sampling units (samples or groups) that have been extracted from a lot with the process ensuring all
sampling units of the lots have an equal probability of being selected and not altered in any way that
would change the analytical result
Note 1 to entry: The extraction process can be a multi-stage process.
3.16
reject/accept criterion
maximum number of deviant seed/grain groups (3.12) that can be detected in the test sample (3.20) of an
acceptable seed/grain lot (3.18)
3.17
seed/grain group
group
determined number of seeds/grains prepared from a seed/grain test sample (3.20) by representative
sampling
3.18
seed/grain lot
lot
population for which sampling is intended to estimate the measured parameter
3.19
stacked event
accumulation of two or more transformation events as a result of traditional breeding and/or successive
transformation steps)
Note 1 to entry: In the context of this document a stacked event refers to a stack in which the two or more events
are not genetically linked.
[SOURCE: ISO 16577:2016, 3.197, modified — Note 1 to entry has been added.]
3.20
test sample
sample prepared for testing or analysis, the whole quantity or part of it being used for testing or
analysis at one time
Note 1 to entry: The test sample is prepared from the laboratory sample (3.9).
Note 2 to entry: The test sample is expected to represent the laboratory sample (3.9).
[SOURCE: ISO 16577:2016, 3.210, modified — Note 1 to entry and Note 2 to entry have been added.]
3.21
testing plan
plan specifying group testing (3.8) conditions including group size (3.7), the number of seed/grain groups
(3.17) and the number of deviant seed/grain groups (3.12) in test sample (3.20) resulting in rejection of
seed/grain lot (3.18)
3.22
threshold
maximum acceptable content of GMO presence in a lot
Note 1 to entry: This can be a prescribed value.
Note 2 to entry: Thresholds can be expressed in mass fraction (3.11) with the proviso that an uncertainty factor
is involved in the conversion to a seed/grain percentage threshold.
4 Principle
4.1 General
In this method, the test sample is divided into a predetermined number of groups. Each group consists
of a determined number of seed/grain and is tested qualitatively for the presence or absence of a GM
target. A statistical evaluation is performed on the number of GM positive groups relative to the total
number of seed/grain groups to determine the GM content in mass fraction.
A statistical calculation determines the optimal testing conditions, namely, the number of seeds/grains
per group (group size), the number of seed/grain groups, and the maximum number of GMO positive
seed/grain groups for seed/grain lot acceptance. Alternatively, a statistical calculation provides an
estimate of the percentage by number of the GM seeds/grains in a lot, according to a given testing plan.
4 © ISO 2021 – All rights reserved
4.2 Preparation of seed/grain groups
Key
1 bulk seed/grain lot
2 laboratory sample
3 test sample
4 seed/grain groups
5 deviant seed/grain
NOTE Each group is represented as an array on the right.
Figure 1 — Sampling illustration of the obtention of seed/grain groups from a bulk seed/grain
lot
The process of forming seed/grain groups from a series of sampling steps starting with the bulk seed/
grain lot is shown in Figure 1, (1).
Although the procedures for obtaining a laboratory sample from a seed/grain lot is not the subject
of this document, a laboratory sample (2) from a seed/grain lot shall be obtained appropriately. The
procedures can be designed according to the References [3], [6], [10], [11], [12], [15], [19] and [23].
The laboratory sample shall be thoroughly mixed and divided/reduced to create the test sample (3).
Likewise, the test sample shall be thoroughly mixed (i.e. homogeneous) and divided into seed/grain
groups (each group represented as an array in Figure 1, (4)) following simple random sampling
principles. The seed/grain groups can vary in size from one single seed/grain up to the complete test
sample (i.e. a single bulk). In most cases, multiple seed/grain groups are created from the test sample.
A determined number of seeds/grains can either be obtained by weighing or a volumetric measurement,
where an approximation of number is made based on a determined conversion factor (e.g. thousand
seeds/grains weight). For the case that weight is used to obtain the seed/grain groups, the operator
shall have an estimate of the variability introduced by using weight rather than seed/grain count.
The group testing procedure described in Clause 7 is carried out on the collective qualitative (positive
or negative) results for each seed/grain group.
4.3 Detection methods for the qualitative analysis of GM seed/grain in seed/grain
groups
[21]
In general, GMO detection methods are categorized into two classes . The first class of assays targets
a nucleic acid sequence for detecting GMO presence. The second class includes methods for detecting a
specified protein that confers a specific transgenic trait. Detection methods from either or both classes
should be selected considering fitness-for-purpose. Guidance on the selection of qualitative methods is
[4]
provided in Clause 8. Further details can be found in ISO 21569 and ISO 21572.
4.4 Statistical evaluation
Sampling and measurement uncertainty shall be considered. Sampling uncertainty can be adequately
[18][2]
considered using the binomial distribution . The FPR and the FNR of the qualitative assay should
[2]
be considered . The LOD of the applied detection method should be considered.
The group testing described here can be used to set reject/accept criteria based on a given threshold by
GMO content, as well as to estimate the GMO content and associated upper and lower confidence limits.
5 Reagents
All reagents used in the analysis should be those specified in the method.
Otherwise, all reagents should be of molecular biology grade.
These reagents shall be stored and used as recommended by the supplier or according to the laboratory
quality assurance specifications. It can also be appropriate to aliquot the reaction solutions required
for the analytical method in order to avoid subjecting them to repeated freeze–thaw cycles, or to reduce
the chances of cross contamination or both. Further details shall refer to ISO 24276 and ISO 21572.
6 Apparatus and equipment
The laboratory should use properly maintained equipment suitable for the methods employed.
Further details shall refer to ISO 24276 and ISO 21572.
7 Design of testing plan
7.1 General
The number of seeds/grains tested, the reject/accept criteria, the sample preparation steps and the
method used for testing shall be determined depending on the analytical purpose.
In seed/grain sample classification, it can be determined whether the number of deviant seeds/grains
or seed/grain groups is above a given reject/accept criterion or not. Then, it can be decided to reject or
accept the seed/grain lot based on the test results.
A basic testing plan for group testing consists of three fundamental parameters:
a) the number of seed/grain groups;
b) the size of the seed/grain groups;
c) the maximum number of deviant seed/grain groups for seed/grain lot acceptance (reject/accept
criterion).
The risks associated with the AQL and the LQL are the producer (alpha) and consumer (beta) risks
respectively, and together with the FPR and FNR allow the design of an appropriate testing plan.
The OC curve can be used to develop a testing plan. Explanations for the estimation of the LOD for a
zero deviant testing plan, the effect of the genome size on the group size if methods targeting DNA are
applied, and the effect of the individual seed size on the sample preparation are given in Annex C.
Annex D provides guidance on the determination of the maximum group size whatever analytical
method is used in the laboratory.
[16]
NOTE Seedcalc is a statistical program (Microsoft Excel spreadsheet application) that is freely available
from the International Seed Testing Association and has procedures to facilitate the design. Seedcalc is located
on the ISTA website.
7.2 Single-stage testing plan
A single-stage testing plan consists of one testing stage. Groups are taken from the test sample and
evaluated once, and a decision is then made based on the results to accept or reject the seed/grain
test sample. In a single-stage testing plan, a specified number of individual seeds/grains or seed/
6 © ISO 2021 – All rights reserved
grain groups shall be selected randomly from the test sample and tested. Depending on the number
of deviants detected and the maximum number of deviants specified in the plan, the seed/grain lot is
either accepted or rejected.
The probability that an individual seed/grain or seed/grain group is deviant, p , can be calculated as
b
given in Formula (1):
m
pP=−11=− 1−p (1)
()
b
where
P is the probability that there are no deviant seeds/grains in the group;
p is the true unknown impurity in the seed/grain lot;
m is the number of individual seeds/grains in a seed/grain group (if seeds/grains are tested indi-
vidually, m = 1).
Then, the probability that a lot will be accepted, P(a) is calculated as given in Formula (2):
c
n
ni−
i
P()a = pp()1− (2)
∑ b b
i
i=0
where
P(a) is the probability that a lot will be accepted;
n is the number of individual seeds/grains or seed/grain groups tested;
c is the maximum number of deviant seed/grain groups for acceptance.
By combining Formulae (1) and (2), P(a) is a function of p, n, m and c.
After n, m and c are determined, an OC curve can be drawn by plotting p and P(a) on the x-axis and
y-axis, respectively.
7.3 Double-stage testing plan
A double-stage testing plan is generally set up so that additional seed/grain groups are tested in the
second stage. Initial seed/grain groups are taken from the test sample and tested. Based on this test
result, three different decisions can be made:
a) accept the seed/grain lot;
b) reject the seed/grain lot; or
c) draw a second set of seed/grain groups from the test sample and retest.
The test results from the first and second stages of testing are combined and used to determine whether
the seed/grain lot is accepted or rejected (see Figure B.1). In Annex B examples for implementation of a
double-stage testing plan to evaluate GMO content in seeds/grains are provided. Subclause B.1 can also
be applied for cases where seed/grain lots are carrying stacked events.
Some additional terms are defined as follows:
— n , the number of independent seed/grain groups to be tested in the first stage;
— n , the number of independent seed/grain groups to be tested in the second stage;
— c , the maximum number of allowable deviant seed/grain groups for acceptance in the first stage;
— c , the minimum number of deviant seed/grain groups that will result in rejection at the first stage;
— c , the maximum number of deviant seed/grain groups in the first and second stages combined
allowed for acceptance;
— d , the number of deviant seed/grain groups in the first stage;
— d , the number of deviant seed/grain groups in the second stage.
P(a) is calculated as given in Formula (3):
n n n
c c −1 ci-
11ni−−ni 2 nj−
1 i 2 i 3 j
11 2
P()a = pp()11− + pp()− × pp()1−
∑ b b ∑ b b ∑ b b
i=0 ic=+1 j=0
i i j
(3)
8 Selection of qualitative methods
8.1 General
An analytical method shall be chosen to meet the purpose of testing. The performance characteristics
of the method should be determined before application in seed/grain testing.
Analytical methods have been developed to detect specific genes encoding transgenic traits or specific
characteristics expressed by specific genes in seeds/grains. Nucleic-acid-based methods such as PCR
[4][5]
are available that detect specific DNA sequences encoding elements, constructs or GMO events .
Protein-based methods such as ELISA and lateral flow immunoassays require a specific antibody for
detecting a specific GM protein (see ISO 21572).
8.2 Performance criteria
The analytical methods applied for the test plan protocol shall detect at least one GM seed/grain in a
group with high probability of detection. Refer to Reference [2].
In the case of PCR, detection methods shall be chosen to meet the purpose of group testing. General
methods performance criteria are described in ISO 24276. General criteria for the design of the testing
plan which should be considered include
a) physical and genome size of seed/grain species as it affects the number of seed/grain that can be
easily ground per group and the number of genome equivalents that can be analysed in a standard
PCR, respectively,
b) absolute PCR limit of detection of the qualitative method, and
c) false-negative rates associated with the method of detection or identification in addition for both
[8][34]
nucleic acid- and protein-based methods should be considered .
[4]
Detection-method-specific performance criteria can refer to ISO 24276, ISO 21569 and ISO 21572.
The seed/grain testing plans discussed in this document assume that the seeds/grains tested are a
representative sample drawn from the seed/grain test sample. Simple representative sampling implies
that each seed/grain in the test sample has both an equal and an independent chance of being included
in the seed/grain group.
9 Interpretation
In determining whether to “accept” or “reject” a given seed/grain lot, the test results shall be compared
with the predetermined reject/accept criterion, e.g. the maximum number of GM-positive groups
allowable for acceptance.
8 © ISO 2021 – All rights reserved
Statistical calculation using the formulae shown below permit the evaluation of a GMO content with
1)
confidence intervals from the test results. Statistical calculation programs such as Seedcalc facilitate
the calculation. In this manner, one can obtain quantitative information on the GMO content of the seed/
grain lot based on how many groups proved to be GM-positive in the qualitative analysis. Together, test
results and their statistical evaluation reveal the level of impurity in the seed/grain lot. Ninety-five
percent upper and lower confidence limits for this impurity evaluation can then be calculated. The true
impurity in the seed/grain test sample can be expected with 95 % confidence to fall within these limits.
The most likely value of GMO content, p, can be evaluated from the test results as given in Formula (4).
d
m
p=−11− (4)
n
where
n is the number of individual seeds/grains or seed/grain groups tested;
m is the number of individual seeds/grains in a seed/grain group (if the seeds/grains are tested
individually, m = 1);
d is the number of deviant seeds/grains or seed/grain groups.
The group testing approach, like quantitative methods, has limitations concerning the GM levels that
can be estimated. Table 1 gives two examples of the highest computed GM estimate for test sample sizes
of 200 seeds/grains and 3 000 seeds/grains. These highest estimates are obtained when all, but one
group is positive. Associated 95 % confidence limits are given to the estimates to show the sampling
uncertainty.
Note For seed/grain group sizes greater than one, when all groups are positive for GM presence, there is
very limited utility in this approach.
Table 1 — Examples of highest computed GM estimate of the content of the deviant seeds/grains
for various seed/grain group sizes (when all but one group is positive) and the 95 % confidence
limits (when all but one group is positive)
Seeds per GM positive Estimated percent- Range of GMO content (%)
Seeds (total) Groups
groups groups age GM seed (for 95 % confidence level)
1 200 0 0,0 0,0 to 1,8
5 40 4 3,9 0,8 to 12,4
10 20 9 10,9 4,0 to 25,8
20 10 19 25,9 13,0 to 48,7
1 3 000 0 0,0 0,0 to 0,1
5 600 4 0,3 0,1 to 0,9
10 300 9 0,8 0,3 to 2,0
3 000
20 150 19 2,0 0,9 to 4,4
30 100 29 3,3 1,7 to 6,8
60 50 59 7,9 4,7 to 14,4
1) Seedcalc is an example of a statistical tool for seed testing. This information is given for the convenience of users
of this document and does not constitute an endorsement by ISO of this product.
If the confidence level for evaluation is set at x %, the upper confidence limit of GMO content in the
evaluation can be calculated using the following Formulae (5) and (6):
x
α =−1 (5)
where x is the confidence level in percentage terms.
()dF+1 m
12−+α ,,dn22 −2d
P =−11− (6)
UL
()nd− ++()dF1
12−+α ,,dn22 −2d
where the quantity F is the 1 − α quantile from an F-distribution with 2d + 2 and 2n − 2d
12−+α ,,dn22 −2d
degrees of freedom.
Also, the two-sided confidence interval (upper limit, P ; lower limit, P ) can be calculated using the
UL LL
following Formulae (7) and (8):
dF+1
() m
12−+α/,22dn,22− d
P =−11− (7)
UL
nd− ++dF1
() ()
12−+α/,22dn,22− d
m
d
P =−11− (8)
LL
dn+−()dF+1 /
α/,22dn,22−+d 2
10 Expression of results
10.1 Classification of a seed/grain lot into “accept” or “reject” category
To classify a seed/grain lot into the “accept” or “reject” category, a statement can be made such that the
seed/grain lot is acceptable or that the seed/grain lot should be rejected.
The upper 95 % confidence limit of the concentration based on the result can be included, or the number
of groups tested, or the number of deviant pools or all of these.
The OC curve expressing the characteristic of sampling can be attached along with the alternative
decision result in order to facilitate understanding.
10.2 Estimation of the level of molecular biomarker in the seed/grain lot
The GMO content in the seed/grain lot can be estimated as described in Clause 9. A statement can be
made such that the most probable value of GMO content is p %, and the ()1−×α 100 % confidence
interval ranges from P % to P %.
LL UL
11 Test report
The test report shall be written in accordance with ISO 24276 and shall contain at least the following
additional information:
a) the sample;
b) a reference to the method that was used for the extraction of nucleic acid or protein;
c) a reference to the methods used for the amplification of the nucleic acid target sequences or the
methods used for the detection of the target protein or both;
10 © ISO 2021 – All rights reserved
d) the LOD of the method used to test the groups and the matrix used to identify the LOD;
e) the reference material used if applicable;
f) the results expressed according to Clause 10;
g) the International Standard used (i.e. ISO 22753:2021);
h) any deviations from the procedure;
i) any unusual features observed;
j) the date of the test.
Annex A
(informative)
Terms and definitions comparison table
A.1 Comparison of terms defined other documents
Synonymous terms defined in other documents or organizations are shown in Table A.1.
Table A.1 — Terms comparison table
a,b c
This document ISTA JRC
Term Definition Term Definition Term Definition
a lot is a distinct
and specified
quantity of ma-
population for which a seed lot is a specified
terial dispatched
seed/grain sampling is intended to quantity of seed that is
seed lot lot or received at one
lot estimate the measured physically and uniquely
time and covered
parameter identifiable
by a particular
contract or ship-
ping document
a submitted sample is
a sample that is to be
submitted to the testing
sample or subsample(s)
laboratory and may com- sample as pre-
received by the labo-
prise either the whole of pared (form the
ratory
the composite sample or lot) for sending
laboratory submitted laboratory
Note 1 to entry: The
a subsample thereof. The to the laboratory
sample sample sample
seed/grain sample
submitted sample may be and intended for
received is expected
divided into subsamples inspection or
to represent the seed/
packed in different ma- testing
grain lot.
terial meeting conditions
for specific tests (e.g.
moisture or health).
a
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 2: Sampling. International Rules for Seed Testing 2021, 2021,
[15]
Bassersdorf, Switzerland .
b
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 19: Testing for seeds of genetically modified organisms.
[17]
International Rules for Seed Testing 2021, 2021, Bassersdorf, Switzerland .
c
EUROPEAN COMMISSION, JOINT RESEARCH CENTRE. (2014), JRC Technical Report: Guidelines for sample preparation
[9]
procedures in GMO analysis .
12 © ISO 2021 – All rights reserved
Table A.1 (continued)
a,b c
This document ISTA JRC
Term Definition Term Definition Term Definition
sample prepared for
testing or analysis,
the whole quantity or the working sample is the
part of it being used for whole of the submitted
testing or analysis at sample or a subsam- (sub-)sample
one time ple thereof, on which prepared from
one of the quality tests the laborato-
Note 1 to entry: The working
test sample described in these ISTA test sample ry sample and
test sample is prepared sample
Rules is made and must from which test
from the laboratory
be at least the weight portions will be
sample.
prescribed by the ISTA taken
Note 2 to entry: The Rules for the particular
test sample is expected test.
to represent the labo-
ratory sample.
A seed group is one of the
portions of the working
sample that is separately
determined number of
prepared (e.g. grinding,
seeds/grains prepared
seed/grain DNA or protein extrac-
from a seed/grain test seed group - -
group tion) and analysed (e.g.
sample by representa-
end-point PCR, ELISA,
tive sampling
real-time PCR) when
using the group testing
approach
a
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 2: Sampling. International Rules for Seed Testing 2021, 2021,
[15]
Bassersdorf, Switzerland .
b
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 19: Testing for seeds of genetically modified organisms.
[17]
International Rules for Seed Testing 2021, 2021, Bassersdorf, Switzerland .
c
EUROPEAN COMMISSION, JOINT RESEARCH CENTRE. (2014), JRC Technical Report: Guidelines for sample preparation
[9]
procedures in GMO analysis .
Annex B
(informative)
Implementation of the method to evaluate GMO content in seeds/
grains example
B.1 Example 1: Group testing to evaluate GMO content in maize grains
B.1.1 General
This annex provides an example to evaluate GMO content in grains using a double-stage testing plan.
The double stage testing plan is used for checking the appropriateness of food labelling in Japan.
B.1.2 Analytical purpose
If there is commingling of stacked event(s) into seed/grain lots, GMO amounts measured by real-time
PCR lead to an overestimation as compared to the actual proportional GM amount in a lot. The group
testing strategy was introduced to estimate the GM content towards the given reject/accept criterion,
irrespective of the presence of stacked event seeds/grains in the lot.
The specified sampling strategy was devised to determine if the GMO content in maize grain lots
exceeds 5 % (mass/mass) or not.
B.1.3 Properties of the analytical method
[19]
Properties for each item are shown in Table B.1. Flow chart for decision making is shown in Figure B.1.
[22][13]
Table B.1 — Analytical method example for double stage sampling plan
Item Value
Type of analytical sample Maize grain
Analyte GMOs in maize grains
Real-time PCR method targeting either or both of Cauliflower
mosaic virus derived 35S promoter and Agrobacterium tumefa-
Qualitative molecular biomarker detection
ciens derived NOS (nopaline synthase) terminator GM elements,
method
which is validated to be able to detect at least 1 GM grain in
20 grains
Type of sampling plan Double stage
Group size in the first stage 20 grains
Number of groups in the first stage 10 groups
6 groups
Reject/accept criterion in the first stage
The second stage is carried out if 7 or more groups show positive.
Group size in the second stage 20 grains
Number of groups in the second stage 10 groups
12 groups after combining the numbers of positive group in the
first and second stages.
Reject/accept criterion in the second stage
Sample lot rejected if ≥ 13 groups test positive for either or both
GM targets.
14 © ISO 2021 – All rights reserved
Figure B.1 — Flow chart for decision making
B.2 Example 2: Group testing to evaluate GMO content in seeds
B.2.1 Analytical purpose
[20]
This example describes a validated German official approach for seed testing . It can be applied to
determine whether the GMO content exceeds 0,1 % or not.
B.2.2 Properties of the analytical method
Properties for each item are shown in Table B.2. Flow chart for decision making is shown in Figure B.2.
The Seedcalc based estimations of GM seed content based on the number of GM positive results obtained
[20][14]
in the qualitative tests are given in Tables B.3 and B.4.
Table B.2 — Analytical method example for German official approach
Item Value
Type of analytical sample Seeds (maize; rapeseed)
Analyte Genetically modified seed
Qualitative molecular biomarker detection Real-time PCR method validated to be able to detect at least 1
method genetically modified seed in 3 000 seeds
Type of sampling plan Double stage
Group size in the first stage 1 000 seeds
Number of groups in the first stage 3 groups
When GMO is detected in more than 1 group: GM seed content
exceeds 0,1 %;
Reject/accept criterion in the first stage
second stage is carried out, if GMO is detected only in 1 group.
Group size in the second stage 1 000 seeds
Number of groups in the second stage 3 groups
In additional group(s) GMO detected: GM seed content exceeds
Reject/accept criterion in the second stage
0,1 %
16 © ISO 2021 – All rights reserved
a
Quantitative real time PCR tests or comparison of Cq values can help to distinguish whole GM seeds present
in the sample from positive signals caused by traces of other homogenously present material (e.g. dust, seed
coating).
Figure B.2 — Flow chart for decision making
Table B.3 — Seedcalc estimations of GM seed content based on the number of GM positive
results obtained in the qualitative tests of three groups of 1 000 seeds
Range of GMO
Seeds GM Percentage GM Probability for a
Seeds content (%)
Groups per positive seeds in true GMO content
(total) (for 95 % confi-
group groups laboratory sample of ≤ 0,1 %
a
dence level)
3 000 3 1 000 0 0,00 0,000 0 to 0,122 9 0,950
3 000 3 1 000 1 0,04 0,000 8 to 0,235 8 0,694
3 000 3 1 000 2 0,11 0,009 9 to 0,476 8 0,253
a
The true G
...
NORME ISO
INTERNATIONALE 22753
Première édition
2021-08
Analyse moléculaire de
biomarqueurs — Méthode pour
l'évaluation statistique des résultats
d'analyse obtenus lors des essais
de sous-échantillons multiples de
semences et de graines génétiquement
modifiées — Exigences générales
Molecular biomarker analysis — Method for the statistical evaluation
of analytical results obtained in testing sub-sampled groups of
genetically modified seeds and grains — General requirements
Numéro de référence
DOCUMENT PROTÉGÉ PAR COPYRIGHT
© ISO 2021
Tous droits réservés. Sauf prescription différente ou nécessité dans le contexte de sa mise en œuvre, aucune partie de cette
publication ne peut être reproduite ni utilisée sous quelque forme que ce soit et par aucun procédé, électronique ou mécanique,
y compris la photocopie, ou la diffusion sur l’internet ou sur un intranet, sans autorisation écrite préalable. Une autorisation peut
être demandée à l’ISO à l’adresse ci-après ou au comité membre de l’ISO dans le pays du demandeur.
ISO copyright office
Case postale 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Genève
Tél.: +41 22 749 01 11
E-mail: copyright@iso.org
Web: www.iso.org
Publié en Suisse
ii
Sommaire Page
Avant-propos .iv
Introduction .v
1 Domaine d’application . 1
2 Références normatives .1
3 Termes et définitions . 1
4 Principe. 4
4.1 Généralités . 4
4.2 Préparation des groupes de semences/graines . 5
4.3 Méthodes de détection pour l’analyse qualitative de semences/graines GM dans
des groupes de semences/graines . 5
4.4 Évaluation statistique . 6
5 Réactifs . 6
6 Appareillage et matériel .6
7 Conception du plan d’essai .6
7.1 Généralités . 6
7.2 Plan d’essai à une étape . 7
7.3 Plan d’essai à deux étapes . 7
8 Sélection des méthodes qualitatives . 8
8.1 Généralités . 8
8.2 Critères de performance . 9
9 Interprétation .9
10 Expression des résultats .11
10.1 Classification d’un lot de semences/graines dans la catégorie «accepté» ou «rejeté» . 11
10.2 Estimation de la teneur en biomarqueur moléculaire dans le lot de semences/
graines . 11
11 Rapport d’essai .11
Annexe A (informative) Tableau de comparaison des termes et définitions .12
Annexe B (informative) Exemple d’application de la méthode pour évaluer la teneur
en OGM dans les semences/graines .14
Annexe C (informative) Estimation de la limite de détection pour un plan d’essai conçu
pour détecter les semences/graines GM dans des lots de semences .22
Annexe D (informative) Détermination expérimentale de la taille de groupe maximale .26
Bibliographie .28
iii
Avant-propos
L'ISO (Organisation internationale de normalisation) est une fédération mondiale d'organismes
nationaux de normalisation (comités membres de l'ISO). L'élaboration des Normes internationales est
en général confiée aux comités techniques de l'ISO. Chaque comité membre intéressé par une étude
a le droit de faire partie du comité technique créé à cet effet. Les organisations internationales,
gouvernementales et non gouvernementales, en liaison avec l'ISO participent également aux travaux.
L'ISO collabore étroitement avec la Commission électrotechnique internationale (IEC) en ce qui
concerne la normalisation électrotechnique.
Les procédures utilisées pour élaborer le présent document et celles destinées à sa mise à jour sont
décrites dans les Directives ISO/IEC, Partie 1. Il convient, en particulier, de prendre note des différents
critères d'approbation requis pour les différents types de documents ISO. Le présent document a
été rédigé conformément aux règles de rédaction données dans les Directives ISO/IEC, Partie 2 (voir
www.iso.org/directives).
L'attention est attirée sur le fait que certains des éléments du présent document peuvent faire l'objet de
droits de propriété intellectuelle ou de droits analogues. L'ISO ne saurait être tenue pour responsable
de ne pas avoir identifié de tels droits de propriété et averti de leur existence. Les détails concernant
les références aux droits de propriété intellectuelle ou autres droits analogues identifiés lors de
l'élaboration du document sont indiqués dans l'Introduction et/ou dans la liste des déclarations de
brevets reçues par l'ISO (voir www.iso.org/brevets).
Les appellations commerciales éventuellement mentionnées dans le présent document sont données
pour information, par souci de commodité, à l’intention des utilisateurs et ne sauraient constituer un
engagement.
Pour une explication de la nature volontaire des normes, la signification des termes et expressions
spécifiques de l'ISO liés à l'évaluation de la conformité, ou pour toute information au sujet de l'adhésion
de l'ISO aux principes de l’Organisation mondiale du commerce (OMC) concernant les obstacles
techniques au commerce (OTC), voir www.iso.org/avant-propos.
Le présent document a été élaboré par le comité technique ISO/TC 34, Produits alimentaires, sous-
comité SC 16, Méthodes horizontales pour l'analyse moléculaire de biomarqueurs.
La présente version française de l'ISO 22753:2021 correspond à la version anglaise publiée le 2021-08
et corrigée le 2022-11.
Il convient que l’utilisateur adresse tout retour d’information ou toute question concernant le présent
document à l’organisme national de normalisation de son pays. Une liste exhaustive desdits organismes
se trouve à l'adresse www.iso.org/fr/members.html.
iv
Introduction
Les essais sur les semences et les graines sont utilisés dans le monde entier pour définir commercialement
la pureté des lots de semences et graines.
Les exigences commerciales relatives à l’étiquetage des produits agricoles ayant une teneur en
organismes génétiquement modifiés (OGM) à un seuil spécifié, à la fois sous forme de contaminant pour
les semences/graines et sous forme d’ingrédient alimentaire, se sont généralisées afin de respecter les
réglementations et satisfaire les demandes des consommateurs. Le respect de ces exigences est évalué
à différents stades de la chaîne d’approvisionnement, qui commence souvent par la récolte des graines.
La réaction de polymérisation en chaîne (PCR) quantitative en temps réel peut être utilisée pour
déterminer la teneur en OGM par l’analyse du rapport entre le nombre de copies d’ADN de l’OGM et
le nombre de copies d’ADN spécifique de l’espèce végétale puis la conversion en fraction massique
génétiquement modifiée (GM).
De multiples événements empilés pendant une récolte, notamment ceux générés par le croisement
d’au moins deux événements individuels, sont couramment utilisés dans le domaine de la production
agricole. Les semences ou graines à empilement d’événements contenant de l’ADN d’OGM correspondant
à au moins deux événements GM conjugués ne peuvent pas être différenciées, par une PCR quantitative
seule, des semences multiples au sein du lot contenant chacun un événement GM individuel. Par
conséquent, si l’OGM réel mesuré provient uniquement de semences à empilement d’événements GM,
la teneur en OGM mesurée par PCR quantitative en temps réel d’un seul échantillon entraînera une
surestimation du nombre réel de semences ou graines GM présentes.
La stratégie d'analyse de groupe, décrite dans le présent document, constitue une solution fiable
pour estimer la teneur en OGM en se basant sur le fait que des semences/graines entières constituent
l'échantillon.
Le processus décrit dans le présent document peut fournir une méthode permettant d’estimer avec
précision les pourcentages de semences/graines GM dans un lot, indépendamment de la présence
de semences/graines à empilement d’événements. La teneur en OGM est déterminée pour les sous-
échantillons multiples représentatifs de semences/graines provenant d’un lot, puis analysée d’un point
de vue statistique.
v
NORME INTERNATIONALE ISO 22753:2021(F)
Analyse moléculaire de biomarqueurs — Méthode pour
l'évaluation statistique des résultats d'analyse obtenus
lors des essais de sous-échantillons multiples de semences
et de graines génétiquement modifiées — Exigences
générales
1 Domaine d’application
Le présent document décrit les exigences générales, les modes opératoires et les critères de performance
applicables à l’évaluation de la teneur en semences/graines génétiquement modifiées (GM) dans un lot
par une stratégie d’analyse de groupe qui comprend l’analyse qualitative de sous-échantillons multiples
puis l’évaluation statistique des résultats.
Le présent document est applicable à la stratégie d’analyse de groupe permettant d’estimer la teneur
en OGM sur un pourcentage de semences/graines afin d’en estimer la pureté, d'évaluer si un critère
de rejet/d'acceptation défini est respecté et de déterminer les cas où des lots de semences/graines
contiennent un empilement d’événements.
Le présent document n’est pas applicable aux produits transformés.
NOTE Une description de l’utilisation de la stratégie d’analyse de groupe est donnée dans les Références [1],
[7], [8], [18], [19] et [20].
2 Références normatives
Les documents suivants sont cités dans le texte de sorte qu’ils constituent, pour tout ou partie de leur
contenu, des exigences du présent document. Pour les références datées, seule l’édition citée s’applique.
Pour les références non datées, la dernière édition du document de référence s'applique (y compris les
éventuels amendements).
ISO 16577, Analyse de biomarqueurs moléculaires — Vocabulaire pour les méthodes d’analyse de
biomarqueurs moléculaires dans l’agriculture et la production agroalimentaire
ISO 21572, Produits alimentaires— Analyse des biomarqueurs moléculaires — Méthodes immunochimiques
pour la détection et la quantification des protéines
ISO 24276, Produits alimentaires — Méthodes d'analyse pour la détection des organismes génétiquement
modifiés et des produits dérivés — Exigences générales et définitions
3 Termes et définitions
Pour les besoins du présent document, les termes et définitions donnés dans l’ISO 16577 ainsi que les
suivants s’appliquent.
L’ISO et l’IEC tiennent à jour des bases de données terminologiques destinées à être utilisées en
normalisation, consultables aux adresses suivantes:
— ISO Online browsing platform: disponible à l’adresse https:// www .iso .org/ obp
— IEC Electropedia: disponible à l’adresse https:// www .electropedia .org/
3.1
limite de détection PCR absolue
limite de détection absolue de la réaction de polymérisation en chaîne
LOD PCR absolue
plus petit nombre nominal (moyen) de copies cibles dans le volume de matrice distribué dans chaque
PCR qui offrirait une probabilité acceptable de détection de la cible
3.2
NQA
NQ
A
niveau de qualité acceptable
niveau d’impureté qui est acceptable pour le producteur et que les méthodes de production peuvent
supporter
3.3
risque du consommateur
risque du consommateur (bêta)
probabilité d’acceptation d’un lot à la limite de qualité inférieure (3.10)
3.4
graine/semence déviante
graine/semence considérée comme non conforme en raison de la présence ou de l’absence d’un trait ou
d’une caractéristique particulière
Note 1 à l'article: Pour les besoins du présent document, est considérée comme déviante une semence qui possède
une caractéristique GM inattendue ou imprévue par rapport aux caractéristiques GM connues ou attendues de la
semence/graine.
3.5
taux de faux négatifs
TFN
probabilité qu’un échantillon pour essai (3.20) positif (groupe de semences/graines) connu ait été classé
comme négatif par la méthode
Note 1 à l'article: Le taux de faux négatifs est le nombre de positifs connus mal classés divisé par le nombre total
d’échantillons pour essai (3.20) positifs.
[SOURCE: ISO 16577:2016, 3.63, modifiée — l’abréviation a été ajoutée, «échantillon pour essai positif»
a été remplacé par «échantillon pour essai (groupe de semences/graines) positif» et la formule a été
supprimée.]
3.6
taux de faux positifs
TFP
probabilité qu’un échantillon pour essai (3.20) négatif (groupe de semences/graines) connu ait été classé
comme positif par la méthode
Note 1 à l'article: Le taux de faux positifs est le nombre de négatifs connus mal classés divisé par le nombre total
d’échantillons pour essai (3.20) négatifs.
[SOURCE: ISO 16577:2016, 3.65, modifiée — l’abréviation a été ajoutée, «échantillon pour essai négatif»
a été remplacé par «échantillon pour essai (groupe de semences/graines) négatif» et la formule a été
supprimée.]
3.7
taille du groupe
nombre de semences/graines contenues dans un groupe
3.8
essai de groupe
évaluation statistique de la teneur en analyte à partir de résultats d’analyse qualitative (c’est-à-dire
positifs ou négatifs) pour chaque groupe de semences/graines de l’échantillon pour essai (3.20)
3.9
échantillon pour laboratoire
échantillon ou sous-échantillon(s) reçu(s) par le laboratoire
Note 1 à l'article: Il est attendu que l'échantillon de semence/graine reçu soit représentatif du lot de semences/
graines (3.18).
[SOURCE: ISO 16577:2016, 3.89, modifiée — La Note 1 à l’article a été ajoutée.]
3.10
LQL
L
QL
limite de qualité inférieure
niveau d’impureté le plus élevé qui est acceptable pour le consommateur
Note 1 à l'article: Ce terme peut être équivalent au seuil (3.22).
3.11
fraction massique
rapport entre les semences/graines GM et les semences/graines totales correspondant au rapport de
masse
3.12
nombre de groupes de graines/semences déviantes
nombre de groupes de semences/graines (3.17) comprenant une ou plusieurs semences/graines déviantes
(3.4)
3.13
courbe caractéristique opérationnelle
courbe OC
représentation graphique du pourcentage de semences/graines déviantes et de la probabilité
d’acceptation situés respectivement sur les axes horizontal et vertical et utilisée dans le cadre de
contrôles qualité afin de déterminer la probabilité d’acceptation de lots de semences/graines (3.18) d’un
plan d’essai (3.21)
3.14
risque du producteur
risque du producteur (alpha)
probabilité de rejet d’un lot au NQA (3.2)
3.15
échantillon représentatif
unités d’échantillonnage (échantillons ou groupes) qui ont été extraites d’un lot au moyen d’un processus
garantissant que toutes les unités d’échantillonnage ont les mêmes chances d’être sélectionnées et n’ont
pas été modifiées au point de fausser le résultat d’analyse
Note 1 à l'article: Le processus d’extraction peut être un processus à multiples étapes.
3.16
critère de rejet/d’acceptation
nombre de groupes de semences/graines déviantes (3.12) maximal qui peut être détecté dans l’échantillon
pour essai (3.20) d’un lot de semences/graines (3.18) acceptables
3.17
groupe de semences/graines
groupe
nombre déterminé de semences/graines préparées à partir d’un échantillon pour essai (3.20) de
semences/graines par échantillonnage représentatif
3.18
lot de semences/graines
lot
population destinée à être échantillonnée pour estimer le paramètre mesuré
3.19
empilement d’événements
accumulation d’au moins deux événements de transformation à la suite d’un croisement traditionnel et/
ou d’étapes de transformation génétique successives
Note 1 à l'article: Dans le contexte du présent document, un empilement d'événements désigne un empilement
dans lequel les deux événements ou plus n’ont pas de liens génétiques.
[SOURCE: ISO 16577:2016, 3.197, modifiée — La Note 1 à l’article a été ajoutée.]
3.20
échantillon pour essai
échantillon préparé pour essai ou analyse, la quantité totale ou une partie de celui-ci étant utilisée pour
l’essai ou l’analyse en une seule fois
Note 1 à l'article: L’échantillon pour essai est préparé à partir d’un échantillon pour laboratoire (3.9).
Note 2 à l'article: Il est attendu que l’échantillon pour essai soit représentatif de l’échantillon pour laboratoire
(3.9).
[SOURCE: ISO 16577:2016, 3.210 modifiée — La Note 1 à l'article et la Note 2 à l’article ont été ajoutées.]
3.21
plan d’essai
plan spécifiant les conditions de l’essai de groupe (3.8), y compris la taille du groupe (3.7), le nombre de
groupes de semences/graines (3.17) et le nombre de groupes de semences/graines déviantes (3.12) au sein
de l’échantillon pour essai (3.20) à la suite du rejet d’un lot de semences/graines (3.18)
3.22
seuil
teneur maximale acceptable d’un OGM dans un lot
Note 1 à l'article: Il peut s’agit d’une valeur prescrite.
Note 2 à l'article: Les seuils peuvent être exprimés en fraction massique (3.11), à condition qu’un facteur
d’incertitude soit utilisé pour la conversion en seuil de pourcentage de semences/graines.
4 Principe
4.1 Généralités
Dans cette méthode, l'échantillon pour essai est divisé en un nombre de groupes prédéterminé.
Chaque groupe comprend un nombre déterminé de semences/graines et est analysé qualitativement
par rapport à la présence ou l’absence d’une cible GM. Une évaluation statistique est effectuée sur le
nombre de groupes positifs aux OGM par rapport au nombre total de groupes de semences/graines afin
de déterminer la teneur en OGM en fraction massique.
Un calcul statistique détermine les conditions d’essai optimales, à savoir le nombre de semences/
graines par groupe (taille du groupe), le nombre de groupes de semences/graines et le nombre maximal
de groupes de semences/graines positifs aux OGM pour l’acceptation du lot de semences/graines. Un
calcul statistique fournit également une estimation du pourcentage en nombre de semences/graines
GM d’un lot, en fonction d’un plan d’essai défini.
4.2 Préparation des groupes de semences/graines
Légende
1 lot de semences/graines en vrac
2 échantillon pour laboratoire
3 échantillon pour essai
4 groupes de semences/graines
5 graine/semence déviante
NOTE Chaque groupe est représenté sous la forme d’une rangée à droite.
Figure 1 — Illustration schématique de l’obtention de groupes de semences/graines à partir
d’un lot de semences/graines en vrac
Le processus de formation de groupes de semences/graines à partir d’une série d'étapes
d'échantillonnage, en commençant par le lot de semences/graines en vrac, est illustré à la Figure 1, (1).
Bien que les modes opératoires permettant d’obtenir un échantillon pour laboratoire à partir d’un lot
de semences/graines ne soient pas abordés par le présent document, un échantillon pour laboratoire (1)
provenant d’un lot de semences/graines doit être obtenu de manière appropriée. Les modes opératoires
peuvent être conçus d’après les Références [3], [6], [10], [11], [12], [15], [19] et [23].
L’échantillon pour laboratoire doit être soigneusement mélangé et divisé/réduit pour créer l’échantillon
pour essai (3). De la même manière, l’échantillon pour essai doit être soigneusement mélangé (c’est-à-
dire, homogénéisé) et divisé en groupes de semences/graines [chaque groupe étant représenté sous la
forme d’une rangée à la Figure 1, (4)] en respectant les principes de l’échantillonnage aléatoire simple.
Les groupes de semences/graines peuvent varier en taille (d’une seule semence/graine à un échantillon
pour essai complet, soit un seul volume). Dans la plupart des cas, plusieurs groupes de semences/
graines sont créés à partir de l’échantillon pour essai.
Un nombre déterminé de semences/graines peut être obtenu par pesage ou par mesure de volume,
auquel cas une approximation du nombre est effectuée sur la base d’un facteur de conversion déterminé
(par exemple, poids de mille semences/graines). Dans le cas où ce poids est utilisé pour obtenir les
groupes de semences/graines, l’opérateur doit avoir une estimation de la variabilité introduite par
l’utilisation du poids plutôt que par le dénombrement des semences/graines.
Le mode opératoire de l’essai de groupe, décrit à l’Article 7 est effectué sur les résultats collectifs
qualitatifs (positifs ou négatifs) pour chaque groupe de semences/graines.
4.3 Méthodes de détection pour l’analyse qualitative de semences/graines GM dans
des groupes de semences/graines
[21]
En général, les méthodes de détection d’OGM sont divisées dans deux catégories . La première
catégorie d’analyse cible une séquence d’acide nucléique pour détecter la présence d’OGM. La seconde
fait appel à des méthodes de détection d’une protéine spécifiée qui confère un trait transgénique
spécifique. Il convient de sélectionner les méthodes de détection parmi l’une des catégories ou les deux
catégories, en tenant compte de leur adéquation avec l’objectif. Des recommandations sur la sélection
des méthodes qualitatives sont fournies à l’Article 8. Des informations supplémentaires sont données
[4]
dans l’ISO 21569 et dans l’ISO 21572.
4.4 Évaluation statistique
L’incertitude d’échantillonnage et de mesure doit être prise en compte. L'incertitude d'échantillonnage
[18][2]
peut être adéquatement prise en compte en utilisant la distribution binomiale . Il convient de tenir
[2]
compte du TFP et du TFN de l’analyse qualitative . Il convient de tenir compte de la LOD de la méthode
de détection appliquée.
L’essai de groupe décrit ici peut être utilisé pour définir les critères de rejet/d’acceptation en fonction
d’un seuil défini par la teneur en OGM, ainsi que pour estimer la teneur en OGM et les limites de confiance
supérieure et inférieure associées.
5 Réactifs
Il convient que tous les réactifs utilisés lors de l’analyse soient ceux spécifiés dans la méthode.
En l'absence de spécification, il convient que tous les réactifs soient de qualité analytique reconnue.
Ces réactifs doivent être conservés et utilisés selon les recommandations du fournisseur ou
conformément aux spécifications d’assurance qualité du laboratoire. Il peut également être approprié
d’aliquoter les solutions réactionnelles requises pour la méthode d’analyse afin d’éviter de les soumettre
à des cycles répétés de congélation-décongélation et/ou de réduire les risques de contamination croisée.
Pour plus d’informations, voir l’ISO 24276 et l’ISO 21572.
6 Appareillage et matériel
Il convient que le laboratoire utilise un matériel correctement entretenu et adapté aux méthodes
employées.
Pour plus d’informations, voir l’ISO 24276 et l’ISO 21572.
7 Conception du plan d’essai
7.1 Généralités
Le nombre de semences/graines soumises à essai, les critères de rejet/d’acceptation, les étapes de
préparation des échantillons et la méthode d’analyse utilisée doivent être déterminés en fonction de
l’objectif de l’analyse.
Lors de la classification des échantillons de semences/graines, il est possible de déterminer si le nombre
de semences/graines déviantes ou groupes de semences/graines est supérieur ou non à un critère de
rejet/d’acceptation défini. Il peut ensuite être décidé de rejeter ou d’accepter le lot de semences/graines
d’après les résultats d’essai.
Un plan d’essai de base pour l’essai de groupe comprend trois paramètres fondamentaux:
a) le nombre de groupes de semences/graines;
b) la taille des groupes de semences/graines;
c) le nombre maximal de groupes de graines/semences déviantes pour l’acceptation du lot de
semences/graines (critère de rejet/d’acceptation).
Les risques associés au NQA et à la LQL sont respectivement les risques du producteur (alpha) et du
consommateur (bêta) qui, avec le TFP et le TFN, permettent de concevoir un plan d’essai approprié.
La courbe OC peut être utilisée pour élaborer un plan d’essai. Des explications concernant l’estimation
de la LOD pour un plan d’essai déviant au zéro, l’effet de la taille du génome sur la taille du groupe si des
méthodes de ciblage de l’ADN sont appliquées, et l’effet de la taille de chaque semence sur la préparation
des échantillons sont données à l’Annexe C.
L’Annexe D fournit des recommandations sur la détermination de la taille maximale du groupe, quelle
que soit la méthode d’analyse utilisée au sein du laboratoire.
[16]
NOTE Seedcalc est un logiciel statistique (tableur Microsoft Excel) disponible gratuitement auprès de
l’International Seed Testing Association et intègre des fonctions permettant de faciliter la conception. Seedcalc
est disponible sur le site web de l’ISTA.
7.2 Plan d’essai à une étape
Un plan d’essai à une étape comprend une étape d’essai. Des groupes sont prélevés dans l’échantillon
d’essai et évalués une fois. Une décision est ensuite prise, en fonction des résultats, d’accepter ou de
rejeter l’échantillon pour essai de semence/graine. Dans un plan d’essai à une étape, un nombre spécifié
de semences/graines individuelles ou groupes de semences/graines doit être choisi au hasard dans
l’échantillon pour essai et soumis à essai. Selon le nombre de déviants détectés et le nombre maximal de
déviants spécifiés dans le plan, le lot de semences/graines est soit accepté soit rejeté.
La probabilité pour qu'une semence/graine individuelle ou un groupe de semences/graines soit déviant,
p , peut être calculée comme indiqué dans la Formule (1):
b
m
pP=−11=− 1−p (1)
()
b
où
P est la probabilité pour qu’aucune semence/graine du groupe ne soit déviante;
p est l’impureté inconnue vraie dans le lot de semences/graines;
m est le nombre de semences/graines individuelles dans un groupe de semences/graines (si les
semences/graines sont soumises à essai individuellement, m = 1).
La probabilité pour qu’un lot soit accepté, P(a), est ensuite calculée comme indiqué dans la Formule (2):
c
n
i ni−
P()a = pp()1− (2)
b b
∑
i
i=0
où
P(a) est la probabilité pour qu’un lot soit accepté;
n est le nombre de semences/graines individuelles ou groupes de semences/graines soumis à essai;
c est le nombre maximal de groupes de semences/graines déviantes pour l’acceptation.
En combinant les Formules (1) et (2), P(a) est une fonction de p, n, m et c.
Une fois que n, m et c ont été déterminés, une courbe OC peut être tracée en représentant p et P(a),
respectivement, sur l’axe des abscisses et l’axe des ordonnées.
7.3 Plan d’essai à deux étapes
Un plan d’essai à deux étapes est généralement établi pour pouvoir soumettre à essai des groupes
de semences/graines supplémentaires lors de la seconde étape. Les groupes de semences/graines de
départ sont prélevés dans l'échantillon pour essai et soumis à essai. En fonction de ce résultat d’essai,
trois décisions différentes peuvent être prises:
a) accepter le lot de semences/graines;
b) rejeter le lot de semences/graines; ou
c) prélever un second ensemble de groupes de semences/graines dans l'échantillon pour essai et
recommencer l’essai.
Les résultats d’essai des première et seconde étapes d’essai sont combinés et utilisés pour déterminer
si le lot de semences/graines est accepté ou rejeté (voir Figure B.1). L’Annexe B fournit des exemples
d’application d’un plan d’essai à deux étapes pour évaluer la teneur en OGM dans des semences/
graines. Le paragraphe B.1 peut également être appliquée dans les cas où les lots de semences/graines
contiennent un empilement d'événements.
Certains termes supplémentaires sont définis ci-après:
— n est le nombre de groupes de semences/graines indépendants qui doit être soumis à essai lors de
la première étape;
— n est le nombre de groupes de semences/graines indépendants qui doit être soumis à essai lors de
la seconde étape;
— c est le nombre maximal de groupes de semences/graines déviantes, autorisé pour l’acceptation
lors de la première étape;
— c est le nombre minimal de groupes de semences/graines déviantes qui conduit au rejet lors de la
première étape;
— c est le nombre maximal de groupes de semences/graines déviantes lors des première et seconde
étapes combinées, autorisé pour l’acceptation;
— d est le nombre de groupes de semences/graines déviantes lors de la première étape;
— d est le nombre de groupes de semences/graines déviantes lors de la seconde étape.
P(a) est calculée comme indiqué dans la Formule (3):
n n n
c c −1 ci-
11ni−−ni 2 nj−
1 i 2 i 3 j
11 2
P()a = pp()11− + pp()− × pp()1−
∑ b b ∑ b b ∑ b b
i=0 ic=+1 j=0
i i j
(3)
8 Sélection des méthodes qualitatives
8.1 Généralités
Une méthode d’analyse doit être choisie pour répondre à l’objectif de l’essai. Il convient de déterminer
les caractéristiques de performance de la méthode avant de soumettre à essai les semences/graines.
Des méthodes d’analyse ont été développées pour détecter des gènes spécifiques codant pour des
traits transgéniques ou des caractéristiques spécifiques exprimés par des gènes spécifiques dans les
semences/graines. Les méthodes basées sur les acides nucléiques, telles que la PCR, permettent de
[4]
détecter des séquences d’ADN spécifiques codant pour des éléments, constructions ou événements GM.
[5]
Les méthodes basées sur les protéines, telles que la méthode ELISA et l’immunochromatographie,
nécessitent un anticorps spécifique pour détecter une protéine GM spécifique (voir l’ISO 21572).
8.2 Critères de performance
Les méthodes d’analyse appliquées pour le protocole du plan d’essai doivent détecter au moins une
semence/graine GM dans un groupe avec une haute probabilité de détection. Voir la Référence [2].
Dans le cas d’une PCR, les méthodes de détection doivent être choisies pour répondre à l’objectif de
l’essai de groupe. Les critères généraux de performance des méthodes sont décrits dans l’ISO 24276. Les
critères généraux de conception du plan d’essai qu’il convient de prendre en compte sont les suivants:
a) taille physique et taille du génome de l’espèce de semence/graine, dans la mesure où elle affecte
le nombre de semences/graines qui peuvent être facilement broyées par groupe et le nombre
d’équivalents génomiques qui peuvent être analysés lors d’une PCR standard, respectivement;
b) limite de détection PCR absolue de la méthode qualitative; et
c) taux de faux négatifs associés à la méthode de détection ou d’identification en plus des méthodes
[8][34]
basées sur les acides nucléiques et les protéines .
Les critères de performance spécifiques de la méthode de détection peuvent figurer dans l’ISO 24276,
[4]
ISO 21569 et l’ISO 21572.
Les plans d’essai des semences/graines décrits dans le présent document supposent que les semences/
graines soumises à essai sont un échantillon représentatif prélevé dans l’échantillon pour essai de
semence/graine. Un échantillonnage représentatif simple implique que chaque semence/graine
présente dans l’échantillon pour essai a une chance égale et indépendante d’être incluse dans le groupe
de semences/graines.
9 Interprétation
Lors de la détermination consistant à «accepter» ou «rejeter» un lot de semences/graines particulier,
les résultats d’essai doivent être comparés avec le critère de rejet/d’acceptation prédéterminé, par
exemple le nombre maximal de groupes positifs aux OGM, autorisé pour l’acceptation.
Le calcul statistique utilisant les formules indiquées ci-après permet d'évaluer une teneur en OGM avec
les intervalles de confiance extraits des résultats d’essai. Des logiciels de calcul statistique tels que
1)
Seedcalc facilitent le calcul. De cette manière, il est possible d’obtenir des informations quantitatives
sur la teneur en OGM du lot de semences/graines à partir du nombre de groupes jugés positifs aux
OGM lors de l’analyse qualitative. Une fois combinés, les résultats d’essai et leur évaluation statistique
révèlent le niveau d’impureté du lot de semences/graines. Des limites de confiance supérieure et
inférieure à 95 % peuvent ensuite être calculées pour cette évaluation du niveau d’impureté. Il est
possible de prévoir avec une confiance de 95 % que le niveau d’impureté vrai dans l’échantillon pour
essai de semence/graine se situe dans ces limites.
La valeur la plus probable de la teneur en OGM, p, peut être évaluée d’après les résultats d’essai, comme
indiqué dans la Formule (4).
d
m
p=−11− (4)
n
où
n est le nombre de semences/graines individuelles ou groupes de semences/graines soumis à essai;
1) Seedcalc est un exemple de logiciel statistique pour l’essai des semences. Cette information est donnée
à l'intention des utilisateurs du présent document et ne signifie nullement que l'ISO approuve ou recommande
l'emploi exclusif du produit ainsi désigné.
m est le nombre de semences/graines individuelles dans un groupe de semences/graines (si les
semences/graines sont soumises à essai individuellement, m = 1);
d est le nombre de semences/graines déviantes ou groupes de semences/graines déviantes.
L’approche de l’essai de groupe, telle que les méthodes quantitatives, est limitée au niveau des teneurs en
OGM qui peuvent être estimées. Le Tableau 1 donne deux exemples d’estimations de teneurs maximales
en OGM pour des tailles d'échantillons pour essai de 200 semences/graines et 3 000 semences/graines.
Ces estimations de teneurs maximales sont obtenues lorsque tous les groupes, sauf un, sont positifs.
Des limites de confiance à 95 % associées sont données aux estimations pour illustrer l’incertitude de
l’échantillonnage.
NOTE Pour les tailles de groupes de semences/graines supérieures à 1, lorsque tous les groupes sont positifs
aux OGM, cette approche est très limitée.
Tableau 1 — Exemples d’estimations des teneurs maximales en OGM dans les semences/graines
déviantes pour différentes tailles de groupes de semences/graines (lorsque tous les groupes,
sauf un, sont positifs) et de limites de confiance à 95 % (lorsque tous les groupes, sauf un,
sont positifs)
Groupes Gamme de teneur en OGM
Semences Graines par Pourcentage estimé
Groupes positifs (%) (pour un niveau de
(total) groupes de graines GM
aux OGM confiance de 95 %)
1 200 0 0,0 0,0 à 1,8
5 40 4 3,9 0,8 à 12,4
10 20 9 10,9 4,0 à 25,8
20 10 19 25,9 13,0 à 48,7
1 3 000 0 0,0 0,0 à 0,1
5 600 4 0,3 0,1 à 0,9
10 300 9 0,8 0,3 à 2,0
3 000
20 150 19 2,0 0,9 à 4,4
30 100 29 3,3 1,7 à 6,8
60 50 59 7,9 4,7 à 14,4
Si le niveau de confiance de l’évaluation est égal à x %, la limite de confiance supérieure de la teneur en
OGM lors de l'évaluation peut être calculée à l’aide des Formules (5) et (6) suivantes:
x
α =−1 (5)
où x est le niveau de confiance, en pourcentage.
()dF+1
m
12−+α ,,dn22 −2d
P =−11− (6)
UL
nd− ++dF1
() ()
12−+α ,,dn22 −2d
où la grandeur F est le quantile 1 − α d’une distribution F avec 2d + 2 et 2n − 2d degrés de
12−+α ,,dn22 −2d
liberté.
De plus, l’intervalle de confiance bilatéral (limite supérieure, P ; limite inférieure, P ) peut être
UL LL
calculé à l’aide des Formules (7) et (8) suivantes:
()dF+1
m
12−+α/,22dn,22− d
P =−11− (7)
UL
()nd− ++()dF1
12−+α/,22dn,22− d
m
d
P =−11− (8)
LL
dn+−dF+1 /
()
α/,22dn,22−+d 2
10 Expression des résultats
10.1 Classification d’un lot de semences/graines dans la catégorie «accepté» ou «rejeté»
Pour classer un lot de semences/graines dans la catégorie «accepté» ou «rejeté», il est possible de
mentionner que le lot de semences/graines est acceptable ou qu’il convient de le rejeter.
La limite de confiance supérieure à 95 % de la concentration en fonction du résultat peut être incluse,
et/ou le nombre de groupes soumis à essai, et/ou le nombre de pools déviants.
La courbe OC exprimant la caractéristique de l'échantillonnage peut être jointe au résultat de la décision
alternative, afin de faciliter la compréhension.
10.2 Estimation de la teneur en biomarqueur moléculaire dans le lot de semences/
graines
La teneur en OGM dans le lot de semences/graines peut être estimée comme décrit à l’Article 9. Il est
possible de mentionner que la valeur la plus probable de la teneur en OGM est p %, et que l’intervalle de
confiance à ()1−×α 100 % varie entre P % et P %.
LL UL
11 Rapport d’essai
Le rapport d’essai doit être rédigé conformément à l’ISO 24276 et doit contenir au moins les informations
supplémentaires suivantes:
a) l'échantillon;
b) une référence à la méthode utilisée pour extraire l’acide nucléique ou la protéine;
c) une référence à la méthode utilisée pour amplifier les séquences cibles d’acides nucléiques et/ou
aux méthodes utilisées pour détecter la protéine cible;
d) la LOD de la méthode utilisée pour soumettre à essai les groupes et la matrice utilisée pour
identifier la LOD;
e) le matériau de référence utilisé, le cas échéant;
f) les résultats exprimés conformément à l’Article 10;
g) la Norme internationale utilisée (c’est-à-dire l’ISO 22753:2021);
h) tout écart par rapport au mode opératoire;
i) tout détail inhabituel observé;
j) la date de l’essai.
Annexe A
(informative)
Tableau de comparaison des termes et définitions
A.1 Comparaison des termes définis dans d’autres documents
Des termes synonymes définis dans d’autres documents ou d’autres organismes sont illustrés dans le
Tableau A.1.
Tableau A.1 — Tableau de comparaison des termes
a,b c
Le présent document ISTA JRC
Terme Définition Terme Définition Terme Définition
quantité distincte et
précise de matériau
population destinée
lot de quantité de semences expédié ou reçu en une
à être échantillonnée lot de
semences/ précise, physiquement lot seule fois et couvert
pour estimer le para- semences
graines identifiable par un contrat ou un
mètre mesuré
document d'expédition
particulier
échantillon des-
tiné à être envoyé au
laboratoire d’essai et
pouvant comprendre
l'ensemble de l’échan-
échantillon ou sous-
tillon composite ou un
échantillon(s) reçu(s)
sous-échantillon de
par le laboratoire.
échantillon tel que
celui-ci. L'échantillon
échantillon échantillon préparé (à partir du
Note 1 à l’article: Il est
échantillon soumis peut être divisé
pour pour lot) pour envoi au
attendu que l'échantil-
soumis en sous-échantillons
laboratoire laboratoire laboratoire et destiné à
lon de semence/graine
conditionnés dans
l'inspection ou à l'essai
reçu soit représentatif
différents matériaux
du lot de semences/
conformes aux condi-
graines.
tions applicables aux
essais spécifiques (par
exemple, conditions
hygrométriques ou
sanitaires).
a
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 2: Sampling. International Rules for Seed Testing 2021, 2021,
[15]
Bassersdorf, Suisse .
b
INTERNATIONAL SEED TESTING ASSOCIATION, Chapter 19: Testing for seeds of genetically modified organisms.
[17]
International Rules for Seed Testing 2021, 2021, Bassersdorf, Suisse .
c
COMMISSION EUROPÉENNE, CENTRE COMMUN DE RECHERCHE. (2014), JRC Technical Report: Guidelines for sample
[9]
preparation procedures in GMO analysis .
TTaabblleeaau Au A.1 1 ((ssuuiitte)e)
a,b c
Le présent document ISTA JRC
Terme Définition Terme Définition Terme Définition
échantillon préparé
pour essai ou analyse,
la quantité totale ou
une partie de celui-ci
étant utilisée pour ensemble de l’échan-
l’essai ou l’analyse en tillon soumis ou
une seule fois. sous-échantillon de
(sous-)échantillon
celu
...












Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.
Loading comments...