I Need A List Of Bayesian Snp/Genotype Callers

Question

[Biomed]

Hi,

I need the first published method for Bayesian SNP/genotype calling and a list of all Bayesian SNP callers for next gen sequencing? It would also be hugely beneficial if there is a comparison paper or web site for these methods. Does any one have a handy list like that?

Thank you

Answer 1

[Marina Manrique]

GigaBayes http://bioinformatics.bc.edu/marthlab/GigaBayes

[Brad Chapman]

FreeBayes is the successor to GigaBayes and under active development: http://bioinformatics.bc.edu/marthlab/FreeBayes

[Marina Manrique]

Thanks Brad! It's cool that it operates on BAM files. And it seems they've moved to a MIT license

Answer 2

[Ian]

Two i can immediately think of:

BAMBINO: http://www.ncbi.nlm.nih.gov/pubmed/21278191 https://cgwb.nci.nih.gov/goldenPath/bamview/documentation/index.html

ComB: http://code.google.com/p/comb/. This requires read mapping using PerM and is SOLiD read based.

[lh3]

Comb is good in theory, but as it has not been published, though we do not know its practical performance.

[lh3]

Comb is good in theory, but as it has not been published, so we do not know its practical performance

Answer 3

[lh3]

Gabor's polybayes is the earliest I am aware of, although it is possible that there are earlier work. As to others:

  Atlas-SNP2 & Bayesian & Single & \citet{Shen:2010zr}\\
  CRISP & Testing$^2$ & Pooled$^3$ & \citet{Bansal:2010kx}\\
  Dindel$^4$ & HMM$^5$ & Pooled$^3$ & \citet{dindel}\\
  FreeBayes$^4$ & Bayesian & Multiple & http://github.com/ekg/freebayes\\
  GATK$^4$ & Bayesian & Multiple & \citet{McKenna:2010vn}\\
  glfMultiples$^4$ & Bayesian & Multiple & http://tinyurl.com/glfMultiples\\
  iCORN & Counting & Single & \citet{Otto:2010ly}\\
  MAQ & Bayesian & Single & \citet{Li:2008zr}\\
  QCALL$^4$ & Bayesian & Multiple & \citet{Le:2010bd} \\
  SAMtools$^4$ & Bayesian & Multiple & \citet{Li:2009ys}\\
  SeqEM & Bayesian & Multiple & \citet{Martin:2010dz}\\
  SliderII & Counting & Single & \citet{Malhis:2010qa}\\
  SNP-o-matic & Counting & Single & \citet{Manske:2009ve}\\
  SNPSeeker & LDT$^6$ & Pooled & \citet{Druley:2009vn}\\
  SNVMix2$^4$ & Bayesian & Single & \citet{Goya:2010mi}\\
  SOAPsnp & Bayesian & Single & \citet{Li:2009fu} \\
  SPLINTER & LDT$^6$ & Pooled & \citet{Vallania:2010ys}\\
  Syzygy$^4$ & Bayesian & Pooled$^3$ & http://tinyurl.com/syzygy-snp \\
  VARiD$^4$ & HMM$^3$ & Single & \citet{Dalca:2010rq} \\
  VarScan$^4$ & Counting & Pair$^6$ & \citet{Koboldt:2009fu}\\

[Biomed]

Thanks Heng. A very helpful list as usual. Do you think any of those stand out from the list as using a unique/radically different approach or unusual implementation etc.

[Erik Garrison]

Heng, FreeBayes is also capable of pooled detection.

Answer 4

[Brad Chapman]

• samtools/maq
• GATK's Unified Genotyper

[Biomed]

Hi Brad, Do you know if Maq was the first published method to use bayesian genotype determination for next gen sequencing? Thanks

[Brad Chapman]

Heh, was going to say that Heng would know and it looks like he does. Deferring to that awesome answer.

Answer 5

[bgulko2]

We just released the BSNP Bayesian Genotype caller. It's been tested on Illumina, 454, SOLiD and Sanger human alignments ans has some technology specific bias correction. It requires a samtools pileup as input, but is fully Bayesian, considers both alignment and sequencing quality and doesn't bias towards the reference. If its helpful, have a look at:BSNP/GPhosCS Home Page