Estimating error models for whole genome sequencing using mixtures of Dirichlet-multinomial distributions

Steven H. Wu, Rachel S. Schwartz, David J. Winter, Don Conrad, Reed A. Cartwright

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Motivation: Accurate identification of genotypes is an essential part of the analysis of genomic data, including in identification of sequence polymorphisms, linking mutations with disease and determining mutation rates. Biological and technical processes that adversely affect genotyping include copy-number-variation, paralogous sequences, library preparation, sequencing error and reference-mapping biases, among others. Results: We modeled the read depth for all data as a mixture of Dirichlet-multinomial distributions, resulting in significant improvements over previously used models. In most cases the best model was comprised of two distributions. The major-component distribution is similar to a binomial distribution with low error and low reference bias. The minor-component distribution is overdispersed with higher error and reference bias. We also found that sites fitting the minor component are enriched for copy number variants and low complexity regions, which can produce erroneous genotype calls. By removing sites that do not fit the major component, we can improve the accuracy of genotype calls.

Original languageEnglish (US)
Pages (from-to)2322-2329
Number of pages8
JournalBioinformatics
Volume33
Issue number15
DOIs
StatePublished - Aug 1 2017
Externally publishedYes

Fingerprint

Dirichlet Distribution
Multinomial Distribution
Error Model
Sequencing
Genome
Genes
Genotype
Binomial Distribution
Biological Phenomena
Minor
Mutation
Mutation Rate
Polymorphism
Libraries
Binomial distribution
Low Complexity
Linking
Genomics
Preparation
Model

ASJC Scopus subject areas

  • Statistics and Probability
  • Biochemistry
  • Molecular Biology
  • Computer Science Applications
  • Computational Theory and Mathematics
  • Computational Mathematics

Cite this

Estimating error models for whole genome sequencing using mixtures of Dirichlet-multinomial distributions. / Wu, Steven H.; Schwartz, Rachel S.; Winter, David J.; Conrad, Don; Cartwright, Reed A.

In: Bioinformatics, Vol. 33, No. 15, 01.08.2017, p. 2322-2329.

Research output: Contribution to journalArticle

Wu, Steven H. ; Schwartz, Rachel S. ; Winter, David J. ; Conrad, Don ; Cartwright, Reed A. / Estimating error models for whole genome sequencing using mixtures of Dirichlet-multinomial distributions. In: Bioinformatics. 2017 ; Vol. 33, No. 15. pp. 2322-2329.
@article{455f07edc2b04e62b5c36bc1b1fd27ae,
title = "Estimating error models for whole genome sequencing using mixtures of Dirichlet-multinomial distributions",
abstract = "Motivation: Accurate identification of genotypes is an essential part of the analysis of genomic data, including in identification of sequence polymorphisms, linking mutations with disease and determining mutation rates. Biological and technical processes that adversely affect genotyping include copy-number-variation, paralogous sequences, library preparation, sequencing error and reference-mapping biases, among others. Results: We modeled the read depth for all data as a mixture of Dirichlet-multinomial distributions, resulting in significant improvements over previously used models. In most cases the best model was comprised of two distributions. The major-component distribution is similar to a binomial distribution with low error and low reference bias. The minor-component distribution is overdispersed with higher error and reference bias. We also found that sites fitting the minor component are enriched for copy number variants and low complexity regions, which can produce erroneous genotype calls. By removing sites that do not fit the major component, we can improve the accuracy of genotype calls.",
author = "Wu, {Steven H.} and Schwartz, {Rachel S.} and Winter, {David J.} and Don Conrad and Cartwright, {Reed A.}",
year = "2017",
month = "8",
day = "1",
doi = "10.1093/bioinformatics/btx133",
language = "English (US)",
volume = "33",
pages = "2322--2329",
journal = "Bioinformatics",
issn = "1367-4803",
publisher = "Oxford University Press",
number = "15",

}

TY - JOUR

T1 - Estimating error models for whole genome sequencing using mixtures of Dirichlet-multinomial distributions

AU - Wu, Steven H.

AU - Schwartz, Rachel S.

AU - Winter, David J.

AU - Conrad, Don

AU - Cartwright, Reed A.

PY - 2017/8/1

Y1 - 2017/8/1

N2 - Motivation: Accurate identification of genotypes is an essential part of the analysis of genomic data, including in identification of sequence polymorphisms, linking mutations with disease and determining mutation rates. Biological and technical processes that adversely affect genotyping include copy-number-variation, paralogous sequences, library preparation, sequencing error and reference-mapping biases, among others. Results: We modeled the read depth for all data as a mixture of Dirichlet-multinomial distributions, resulting in significant improvements over previously used models. In most cases the best model was comprised of two distributions. The major-component distribution is similar to a binomial distribution with low error and low reference bias. The minor-component distribution is overdispersed with higher error and reference bias. We also found that sites fitting the minor component are enriched for copy number variants and low complexity regions, which can produce erroneous genotype calls. By removing sites that do not fit the major component, we can improve the accuracy of genotype calls.

AB - Motivation: Accurate identification of genotypes is an essential part of the analysis of genomic data, including in identification of sequence polymorphisms, linking mutations with disease and determining mutation rates. Biological and technical processes that adversely affect genotyping include copy-number-variation, paralogous sequences, library preparation, sequencing error and reference-mapping biases, among others. Results: We modeled the read depth for all data as a mixture of Dirichlet-multinomial distributions, resulting in significant improvements over previously used models. In most cases the best model was comprised of two distributions. The major-component distribution is similar to a binomial distribution with low error and low reference bias. The minor-component distribution is overdispersed with higher error and reference bias. We also found that sites fitting the minor component are enriched for copy number variants and low complexity regions, which can produce erroneous genotype calls. By removing sites that do not fit the major component, we can improve the accuracy of genotype calls.

UR - http://www.scopus.com/inward/record.url?scp=85026346784&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85026346784&partnerID=8YFLogxK

U2 - 10.1093/bioinformatics/btx133

DO - 10.1093/bioinformatics/btx133

M3 - Article

VL - 33

SP - 2322

EP - 2329

JO - Bioinformatics

JF - Bioinformatics

SN - 1367-4803

IS - 15

ER -