A genotype calling algorithm for affymetrix SNP arrays

Abstract

A classification algorithm, based on a multi-chip, multi-SNP approach is proposed for Affymetrix SNP arrays. Current procedures for calling genotypes on SNP arrays process all the features associated with one chip and one SNP at a time. Using a large training sample where the genotype labels are known, we develop a supervised learning algorithm to obtain more accurate classification results on new data. The method we propose, RLMM, is based on a robustly fitted, linear model and uses the Mahalanobis distance for classification. The chip-to-chip non-biological variance is reduced through normalization. This model-based algorithm captures the similarities across genotype groups and probes, as well as across thousands of SNPs for accurate classification. In this paper, we apply RLMM to Affymetrix 100 K SNP array data, present classification results and compare them with genotype calls obtained from the Affymetrix procedure DM, as well as to the publicly available genotype calls from the HapMap project.

Impact of common regulatory single-nucleotide variants on gene expression profiles in whole blood

“…The main differences between BRLMM and the previously-developed RLMM [2,3,4] method lie in the clustering space transformation and in the estimation of cluster centers and variances. There are a number of other potential improvements which have been or are in the process of being evaluated.…”

“…The Dynamic Model (DM, [1]) which has been extensively used for the GeneChip® Human Mapping 100K Array Set and the GeneChip® Human Mapping 500K Array Set has proven to be very effective, however it is possible to do better. Rabbee & Speed recently developed a model called the Robust Linear Model with Mahalanobis distance classifier (RLMM, pronounced 'realm') which provided an improvement over DM on the Mapping 100K set [2,3,4]. We present here an extension of the RLMM model developed for the Mapping 500K product which provides a significant improvement over DM in two important areas -it improves overall performance (call rates and accuracy) and it equalizes the performance on homozygous and heterozygous genotypes.…”

Whole-genome association study of bipolar disorder

“…8 Genotypes were called using the Bayesian Robust Linear Model with Mahalanobis distance classifier. 32 A panel of 24 markers present on the whole genome product as well as 25 single nucleotide polymorphisms (SNPs) previously genotyped in the UCL samples were used as genetic fingerprints to detect sample switches.…”

A regulatory SNP in AKAP13 is associated with blood pressure in Koreans

“…8 Briefly, the accuracy of the genotyping was calculated by Bayesian robust linear modeling using the Mahalanobis distance genotyping algorithm. 15 Samples that had accuracies below 98% and a high missing genotype call rate (X4%), high heterozygosity (430%) or inconsistency in sex were excluded from subsequent analyses. Individuals who had a tumor or were undergoing antihypertensive therapy were excluded.…”