search

spcdot / CopyscAT

Inferring CNV data from single cell ATAC seq
GNU General Public License v3.0
22 stars 6 forks source link

meaning of data in output files #3

Open gigglypear opened 4 years ago

gigglypear commented 4 years ago

Hello again!

I have run your software on both the sample and some of my own. And I am a bit confused about some of the data in the output files:

  1. In clean_cnv, are the cnv scores computed log-transformed?

  2. In the loss output file, what do the -1, 0, 1 values mean?

spcdot commented 4 years ago

1) The CNV score is an expected copy number built from the Z score of the signal. It should be 2 for non-altered chromosomes and closer to 1 and 3 if there is a loss and gain, respectively - if it looks a little off (numbers too high or too close together) you can change the sdCNV parameter from the default of 0.5 (usually something in the 0.3-0.7 range should work). 2) For loss output, -1 is cells with relative loss, 1 is cells with relative gain, and 0 is cells with indeterminate status (that is the calls are too low confidence for it to assign it to either cluster). Even though this method is designed to identify regions of loss, it will also sometimes identify regions of local amplification, so best to look at the data carefully to see which interpretation makes the most sense.

Hope that helps!