EvolBioInf / sid

SNPs from diploid genomes [WIP]

SNPs from diploid genomes

The sid program can be used to detect single nucleotide polymorphisms in aligned sequencing reads. Based on the pileup output of Samtools, for each genome site the most likely genotype is computed.

Installation

To install sid, clone the repository and change into the source directory. To prepare the cloned repository for compilation, cd into the directory and run

./autogen.sh
./configure

The program is then compiled and (optionally) installed to /usr/local/bin with

make
sudo make install

Usage

For preprocessing, a working installation of samtools is needed. Given the aligned sequencing data in SAM or BAM format, a pileup needs to be generated: samtools mpileup input.bam > pileup.dat Useful parameters for the samtools mpileup program are -C 50, which is recommended for BWA-aligned reads, and -q 1 which discards reads with ambiguous mapping. The switches -q and -Q can be used to set a lower threshold for mapping quality and base quality, respectively.

The pileup is then used as the sid input:

sid -m local pileup.dat > output.csv

The -m switch selects the SNP calling method, see sid -h for all available ones.

Output format

The output is given as comma-separated values (CSV).

chrom,pos,label,gt,hom_conf,het_conf,conf_type
1,3003229,hom,GG,2.98235e-13,1,p_value
1,3003230,hom,GG,3.67401e-14,1,p_value
1,3003231,het,GA,1,8.01707e-17,p_value
1,3003232,hom,AA,4.36739e-12,1,p_value

Each line represents the genotype calling result for one genome site, defined by a chromosome (chrom) and a coordinate (pos). The label is either hom or het and indicates a homozygous or heterozygous site, respecitvely, while the gt column lists the called genotype. The next two columns give the confidence in the call as a P-value.