Automated ploidy cutoffs for karyotype assignment in sample_qc #127
Conversation
|
thanks for the review, Laurent -- I made a few more changes based on your comments. let me know what you think! |
|
thanks for the review -- I tried some new changes, let me know what you think! happy to also sit down and chat through this sometime instead in case that's easier |
utils/sample_qc.py
Outdated
| xx_x_stats.mean + (upper_SD_cutoff * xx_x_stats.stdev) | ||
| ] | ||
| cutoffs["Y"] = [ | ||
| abs(xx_y_stats.mean - (upper_SD_cutoff * xx_y_stats.stdev)), |
There was a problem hiding this comment.
I don't think we want abs here. I think max(0, ...) is what we want in theory. Although I don't even think we need this as coverage can never be negative 0 is the same as any negative number in practice.
There was a problem hiding this comment.
Or maybe I misunderstood the purpose of abs here...
There was a problem hiding this comment.
I added this abs because i wanted to use the upper cutoff of Y ploidy in XX samples as the lower cutoff for a single copy of Y (any samples above this cutoff will get marked as at least having one copy of Y). Unfortunately, I realized when testing the code that Y ploidy distribution in the XX samples was so tight that this value was negative (and in this case, max(0, ...) wouldn't work, as every sample would get called as having at least one copy of Y). Happy to change this, let me know what you think!
|
thanks for the suggestions -- the code looks much tidier now! I think we're getting close, let me know what you think about the lower cutoff for single copy of Y (#127 (comment)) |
…X samples (changed '-' to '+')
|
I updated the cutoffs and unnested |
…x_expr (sample_qc.py)
… additional logging statements in get_ploidy_cutoffs
|
Thanks for those suggestions -- I added the edits and an additional set of logging statements (just to send the XX/XY stats and X/Y cutoffs to stdout) |
I decided to have the code check for X and Y for each case, as that made the most sense to me, but I'm open to any feedback!