Error in get_gerp_score_percentile.py and zero gerp scores

[saadarif]

In the following code:

def approx_percentile(p, bin_edges, total):
    # find the 99th percentile from the histogram
    p = 99
 .....

the value ofpis hard coded, so never takes on user supplied values as specified in the docs. I imagine this is not intended?

On a different note, most of the positions in my reference have a gerp score of zero (based on alignment against 101 genomes of outgroup taxa). My current understanding is that Gerp scores of <1 suggest no constraint on the positions. But info on the UCSC genome browser also suggests that gerp scores of 0 might be lack of alignment information.

[verku]

Hi! Thanks for reaching out! You are correct, this was a mistake. I'll publish a corrected script with the next pipeline version release!

It depends how divergent the species in your alignment are, but in general one can expect that many sites will be so divergent that there is not enough alignment information for a GERP score to be calculated and it will be set to zero.