Christopher Hartl 944a7d815e Bringing VQSRV3 up to date. Lots of new features (un-classifying the worst-performing training sites, treating the x% best/worst sites as postive/negative points, ability to pass in a monomorphic track to see ROC curves output). Minor changes to AlleleBalance: weighted average was incorrectly specified (using logscale actually biased the average towards the AB of low-quality genotypes), and breaking out AB by het, hom, and diploid to bring it in line with some (private) changes to the indel likelihood model that (correctly) computes these values for indels.		2012-04-28 11:31:03 -04:00
..
config	Reorganized the codebase beneath top-level public and private directories,	2011-06-28 06:55:19 -04:00
src	Bringing VQSRV3 up to date. Lots of new features (un-classifying the worst-performing training sites, treating the x% best/worst sites as postive/negative points, ability to pass in a monomorphic track to see ROC curves output). Minor changes to AlleleBalance: weighted average was incorrectly specified (using logscale actually biased the average towards the AB of low-quality genotypes), and breaking out AB by het, hom, and diploid to bring it in line with some (private) changes to the indel likelihood model that (correctly) computes these values for indels.	2012-04-28 11:31:03 -04:00
test	Bug fix for: https://getsatisfaction.com/gsa/topics/problem_with_indelrealigner_and_l_unmapped	2012-04-27 09:58:38 -04:00