Entering edit mode
9.8 years ago
reza.jabal
▴
580
Hi,
I'm kind of novice in the field and still trying to absorb fundamental concepts, so please excuse me if my question appears a bit silly! I am wondering if there is a cut-off for heterozygous/homozygous ratio when checking for contamination. Here I am dealing with a logarithmic plot where HET/HOM measures for each samples is drawn against deviation metrics. This has been handed to me from a colleague and I am not fully confident about the interpretation of the plot.
The linked image is not showing. We can't see what you mean. ... EDIT: We can dig into the attempted linked URL and see the figure. It's labeled axes are DEVIATION_METRIC vs HET:HOM. so there's still no way to know what it means. You have to ask your colleague just what is the deviation metric and just what the het vs hom is measuring.
Also, what is the context? Contamination of what? Are these genotype SNP-chips or NGS? Human or Wheat?
Fixed. For future reference Reza, please try to post images to imgur, or somewhere where we don't have to jump through hoops to see what you're asking about!
Guess you just updated it, thanks!
The image seems to not work again.
Hmm. Hopefully fixed.. again!
Karl and Daniel,
Thanks very much for correcting the mess. I am actually working with the exome data of a human stillbirth case, trying to delineate the causative mutation. Having discussed it with the colleague this morning, the plot apparently pertains cross-sample contamination in the batch that has been modeled along with deliberately contaminated samples (dots with <10% contamination). I doubt it she herself have a clear understanding of "deviation_metric"!
This script could give someone nightmares. It would in fact be surprising if anyone actually knew what
deviation_metricmeant.Ok, I got this upon further chase up! but still not sure if it seems reasonable