Methods for high throughput genotyping

• US 8,498,825 B2
• Filed: 06/23/2011
• Issued: 07/30/2013
• Est. Priority Date: 12/07/2005
• Status: Active Grant

• Current Assignee: Affymetrix, Inc. (Thermo Fisher Scientific Incorporated)
• Original Assignee: Affymetrix, Inc. (Thermo Fisher Scientific Incorporated)
• Inventors: Huang, Jing, Jones, Keith W.
• Primary Examiner(s): Brusca, John S
• Application Number:
  US13/167,264
• Publication Number:
  US 20110251798A1
• Time in Patent Office: 768 Days
• Field of Search: None
• US Class Current: 702/20
• CPC Class Codes: C12Q 2600/156 Polymorphic or mutational markers
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 20/20 Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection

Abstract

Methods for genotyping polymorphisms using allele specific probes are disclosed. A training set is used to generate a model for each polymorphism to be interrogated. The training set is used to obtain an estimate of the asymmetry between an intensity measurement for a first allele and an intensity measurement for a second allele of the same polymorphism. The intensity measurement obtained for a test sample is adjusted using the estimate of asymmetry prior to using the intensity measurements to make a genotyping call. In preferred embodiments the adjustment is applied to polymorphisms that have a likelihood of being heterozygous that is above a specified threshold.

Claims

1. A method for genotyping a polymorphism in a test individual with a genotyping array, the method comprising:

(a) obtaining allele specific intensity measurements for a first allele of the polymorphism and for a second allele of the polymorphism in a training set, wherein the training set comprises a plurality of individuals of known genotype for said polymorphism, and wherein the allele specific intensity measurements each represent an intensity of signal associated with one or more features on the genotyping array;

(b) grouping the individuals in the training set into groups according to genotype, wherein each individual is assigned to one of the following groups;

homozygous for the first allele, homozygous for the second allele, or heterozygous;

(c) calculating a summary ratio of the intensity measurements for said first allele to the intensity measurements for said second allele for each group to obtain a first summary ratio for the group that are homozygous for said first allele, a second summary ratio for the group that are homozygous for said second allele and a third summary ratio for the group that are heterozygous;

(d) measuring an allele specific intensity value for said first allele and for said second allele in said test individual with the genotyping array and calculating a test ratio for said test individual;

(e) adjusting the allele specific intensity value for said first allele according to the third summary ratio obtained in (c) if said test ratio is closer to the third summary ratio than to the first or second summary ratios; and

(f) making a genotype call for said polymorphism using the adjusted value obtained in (e) for said first allele and the intensity value obtained in (d) for the second allele.

2. The method of claim 1 wherein the genotype call of step (f) is made using a dynamic modeling algorithm.

3. The method of claim 1 wherein the genotype call of step (f) is made using an expectation-maximization algorithm.

4. A method for calling the genotype of a sample at a selected polymorphism in a sample, the method comprising:

(a) obtaining hybridization intensity values for a first genotyping array for each of a set of training samples, wherein the set of training samples comprises a plurality of training samples of known genotype;

(b) making a genotype call for each of a plurality of SNPs in the each of the training samples using the hybridization intensity values from individual probe quartets;

(c) comparing the genotype call with the known genotype for each probe quartet to identify a plurality of K best probe quartets for each SNP, where K is at least 1, wherein probe quartets are selected as best probe quartets if the genotype call made using said quartet has high concordance with the known genotype for that SNP;

(d) calculating a distribution of (intensity A)/(intensity B) distribution for the training samples for each sub-group of AA, AB and BB to obtain an AA reference distribution, an AB reference distribution and a BB reference distribution;

(e) hybridizing a test sample to a second genotyping array to obtain hybridization intensity values for said K best probe quartets, wherein the second genotyping array is identical to the first genotyping array;

(f) calculating an intensity fraction for each of the plurality of K best probe quartets and comparing the intensity fractions with the reference distributions for AA, AB and BB to determine the likelihood that the polymorphism is AB, wherein the intensity fraction is (intensity A)/(intensity B);

(g) adjusting the intensity of intensity B by the (intensity A)/(intensity B) ratio from the AB group from the reference set to obtain an adjusted allele B intensity if the likelihood that the polymorphism is AB is greater than a selected threshold; and

(h) using the adjusted intensity B value to generate a genotype call using either a dynamic modeling algorithm or an expectation-maximization algorithm.

5. The method of claim 4 wherein K is selected from 2, 3, 4, 5, 6 or 7.

6. The method of claim 5 wherein K is the same for each SNP in the plurality.

7. The method of claim 4 wherein K is selected for each SNP and is the number of quartets for that SNP that predict the correct genotype in the training set with a minimum concordance.

8. A method for determining genotypes, the method comprising:

hybridizing a control sample to a first genotyping array, wherein the genotype of each of a plurality of polymorphisms in the control sample is known, and wherein the genotype comprises at least allele A and allele B;

measuring signal intensity values for each of the plurality of polymorphisms;

making a genotype call from the signal intensity values for each of the plurality of polymorphisms;

identifying whether each genotype call was a correct genotype call;

calculating a ratio of signal intensity values for allele A to signal intensity values for allele B using the correct genotype calls;

hybridizing a test sample to a second genotyping array, which wherein the second genotyping array is identical to the first genotyping array;

normalizing hybridization data from the test sample with the ratio of signal intensity values for allele A to signal intensity values for allele B, thereby more accurately representing allelic frequency; and

determining the genotype of the test sample.

9. The method of claim 8, wherein normalizing comprises multiplying the hybridization data by the ratio of signal intensity values for allele A to signal intensity values for allele B.

10. The method of claim 8, wherein determining comprises identifying each of the plurality of polymorphisms in the test sample.

11. The method of claim 8, wherein determining comprises employing either a dynamic modeling algorithm or an expectation-maximization algorithm.

12. The method of claim 8, wherein both the control sample and the test sample originate from diploid organisms.

13. The method of claim 12, wherein the diploid organisms are humans.