bam2mpg

The program "bam2mpg" calls genotypes from sequence reads of haploid or diploid DNA aligned to a closely-related reference sequence. The MPG (Most Probable Genotype) algorithm is based on a Bayesian model which simulates sampling from one or two alleles with sequencing error, and then calculates the likelihood of each possible genotype given the observed sequence data. Using prior probabilities dependent on the expected heterozygosity of the sequence, MPG then predicts the "Most Probable Genotype" at each site, along with quality scores estimating the accuracy of its calls.

Bam2mpg was written by Nancy F. Hansen, a staff scientist at NHGRI. The current source code and releases for bam2mpg are hosted on github at git://github.com/nhansen/bam2mpg.

Last Modified: Wednesday, 29-Apr-2015 17:08:32 EDT