Accept non-integer ploidy as input to 'call'

[shashwatsahay]

Hi

when calculating purity ploidy using purecn we generally get a non integer ploidy also can we somehow modify cnvkit to accept the non integer ploidy when using the call function

[etal]

Non-integer ploidy? How does PureCN intend that to be used? Does a ploidy of e.g. 3.5 mean the sample is a mix of triploid and quadruploid cells? In that case you could probably round up and get a reasonable result from CNVkit's call command. Alternatively, you could leave the ploidy at 2 and just use the purity input, and CNVkit's results will still be reasonably straightforward to interpret for the common use cases (e.g. was this gene-level loss a partial or full knock-out).

[shashwatsahay]

Well the point being here is that the the physical state of a single tumor cell can be a diploid or tetraploid or hexploid.

But when doing bulk sequencing we aren't mesuring a single tumor cell, we are measuring a mixure of tumor cells which we know can be hetergenous in tumor sampes.

So different tumor clones showing different base ploidy hence a average ploidy will be useful when analysing in bulk.

[etal]

Makes sense, thanks. I'll see what I can do.