search

etal / cnvkit

Copy number variant detection from targeted DNA sequencing
http://cnvkit.readthedocs.org
Other
502 stars 163 forks source link

interpretation of the result #744

Open stroke1989 opened 2 years ago

stroke1989 commented 2 years ago

Hi , I'm a new to CNVKIT, CNVKIT is a excellent tools. I tried to use this tool to analyze my WES data. The following is my plot:

Snipaste_2022-06-29_02-02-03

I have several questions:

  1. what is the value threshold in y-axis that indicates amplification or deletion? as the red row pointed, the loci showed deletion?
  2. what does light grey points and dark grey points mean?
  3. could homozygous deletion OR LOH be inferred from this plot? How? Hope your reponse! Appreciate!
StrawHattM commented 1 year ago

Not the original author, but from my experience:

1.- there's two thresholds, one for the segmentation (in segment call), and one for the scatter plot (in scatter call). If i remember correctly, it's 0.2 in both; but this is just a default and you can set it yourself. call cnvkit.py scatter --help for help, I think the threshold is specified there.

2.- dark gray points are bin-level copy number ratios, they come from the .cnr file, output of fix. Light gray points are segments (from .cns, output of segment/call) that haven't passed the threshold.

3.- the log2 is the log2 fold change from the calculated reference copy number. You can infer deletions and LOH events from it by estimating the difference to the reference. However, best would be to use the cn column in the .cns files that come from the call function, since it's purpose is to infer integer copy numbers.