SNP calling

[brachysclereid]

Reference-based SNP callers work by comparing the sequence aligned to the reference. Are there any tools/piplines that will call SNPs between aligned files? In my case I am interested in calling SNPs between the genotypes I am aligning and not in SNPs present between by genotypes and the reference.

[brachysclereid]

SNP calling between samples....in a polyploid

I should also add that I am trying to call SNPs either without a reference genome or with the diploid version of a polyploid I am working with.

[rfilbert]

Partek Flow calls SNPs either against a reference or between your samples.

[brachysclereid]

Surely there is something free from academic labs that will do this

[rfilbert]

If it makes you feel better, Partek's SNP caller is also about 50x faster than mpileup which is one of the main reasons I went with them. I have not compared it to GATK yet. While it is commercial, it is not expensive and comes with technical support.

[brachysclereid]

Is there a way to call SNPs between samples with samtools mpileup or GATK?

[Zam]

Cortex is designed to call variants between samples, without a reference.

Papers:
1. De novo assembly and genotyping of variants using colored de Bruijn graphs.
Z Iqbal, M Caccamo, I Turner, P Flicek, G McVean, Nature Genetics (2012)

2. High-throughput microbial population genomics using the Cortex variation assembler. Z Iqbal, I Turner, G McVean, Bioinformatics 2012

Software freely available here:

http://cortexassembler.sourceforge.net/index_cortex_var.html

Manual here

http://cortexassembler.sourceforge.net/cortex_var_user_manual.pdf

It can take BAM or fast files as input. Hope that helps!

Zam

[brachysclereid]

SNP calling between samples....in a polyploid

Thanks for all the suggestions. I will check out Cortex. Do you think it can handle polyploids?

[Zam]

Yes, it can call biallelic sites in polyploids. The genotyping models only handle haploid and diploid, so you will have to do discovery only, but the VCF will report coverage on both alleles, so you can post-process to interpret.

[rob123king]

Anyone know how I can do snp calling using a draft reference which is a series of contigs, example head below. I can do it using BWA-MEM and mpileup with a complete reference but using this draft I get almost no data. I think theres an issue with alingment to the draft reference as it is not a series of continuous data, maybe if i use some program to put them all together and fill in the gaps with N's or do the contigs just need renaming? Little lost how could use this draft genome, may just end up comparing against a more distant but complete genome.

>contig:unspecified:411:1:710:1 contig 411
GTTAGTGCCTACTGCCTCTTTTTCATCTGCCTGAATATCTCTCATCTTTCGTTATCCTTA
ATTGTCCTCATATCTTTCGATTATTTGTCAGGTTTTTGAGTATTAGGTGCTTATTCTTTA
AAAAATACACAGAAAGTTTATCGAAATGCTGCATTTAATCGTATTGTATATGAATACGCA
TTCTCTAACTGGTAAATGTATTGTACGAAAGAATTGTGTCTCAATGAAGTAAGTGGTTAT
CTCTGGTGTTAAGTTGGTGAGTATTTGTTGAATATTGATATTAATACCAGCAAATTGGAA
GTGATATTGAGTAAATGTTGGGATAGTAGTAAAGAGAAAGAACTTCGGTGTATAAGCGAA
AACTTCATTTAACCGTTAACAATACGATGGAAATAAGCACATGAAAATAATCTCATTTCC
TAGAATAGAGTCAGATTACTTGCTATTCTTAATTGTTGTTTCAAATGAAGTAGTACTAAC
TACTAACTACAGTAGAAATTGCAATATAATTTCCACTCTCCCTTGTGGAATTGGTAATTT
TGTGAGGGTAATATTAATATGCTTTGCTGCTCCTTAATAAAAATCAAACCATCAACTCTA
CCAAAATTAGTTGGGGTCAGTGATAGGAATTAATATTAGGTTATTTTTTGAAAACATAAT
GTATACACGGAAAGTGAGGAGAAAAAGGACAGGGGAATACAATATGGGAA
>contig:unspecified:586:1:3978:1 contig 586
TATATTTGCCTCATAAGAATGGGCCTATGGGCAAAAATGGGTTGGTCCAGATTGGGCTTT
TATTTATTTACATAGGAAGGGAAAAAGGCCCATAACCCAACAATTTTAACACCACATTCA
TTTACAAATCTACCCATTTGACCGAATCGACTTTTTTTCTTTTCATGTTCTTCTTCTCTC
AAGATAATTTAATTTATGATATTATCACAGTAAAATAAAAATTACTAAAATCTCACGCCG
TTTGCCCATTTTACAAACAAATTATATATCTTGTTGATTTAAATAAAAAAGAGATATATT
TAGATTAATAAATATAAAAGATAACGTAAATTTGTGTCATTTTATAATGAAAGGGGCATA
TATGAATTCAAAATATATTGAAAGGAATGTATTTAGATAAAAAAATATAGCGAAGGGTAT