Obtaining probability of all variant calls being correct

The VCF format specifies quality scores (QUAL) for each variable position (variant) in a genome. The QUAL value is the Phred quality score for the assertion that alternative bases of a variant are correct, that is, \mathrm{QUAL} = -10 \log_{10} p, where p is the probability that the alternative base calls are wrong. Using the QUAL scores, one may easily calculate the probability that all variant calls in a VCF file are correct.

Here we give an equation for that probability, a Python script that implements it and an example of its usage.

Continue reading