README

Bam2mpg

The program "bam2mpg" calls genotypes from sequence reads of haploid or diploid DNA 
aligned to a closely-related reference sequence. The program reads alignments in BAM 
format (http://samtools.sourceforge.net). 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 the calls.

Bam2mpg was written by Nancy Fisher Hansen, a staff scientist in the Genome Technology 
Branch (GTB) of NHGRI, an institute at the National Institutes of Health.  Nancy can 
be reached at nhansen@mail.nih.gov.

If you use bam2mpg in your work, please cite the 
paper: Teer, J.K. et al. 2010. Systematic comparison of three genomic enrichment methods 
for massively parallel DNA sequencing. Genome Research. 20(10):1420-31.

Build Dependencies

* git
* samtools 0.1.16 or above

Build Instructions

Clone the bam2mpg repository:

git clone git://github.com/nhansen/bam2mpg.git

Build bam2mpg:

cd bam2mpg
perl Build.PL
./Build
./Build test
./Build install

* For local installation, call perl Build.PL with "--install_base $HOME"