Strange values for cellular prevalence and variant allele frequency>?

[jmchan88]

Hi there,

I'm using PyClone to estimate cellular prevalence and VAFs across sample timepoints for a single patient, and the resulting cellular prevalence and VAFs seem strange to me. In particular, I have cellular prevalence values that are non-zero when the VAFs are zero. Does this make sense? For example,

In this instance, I had no variant reads calling the ALK mutation until the last sample timepoint -T08-IM6, but there are non-zero cellular prevalences that precede that timepoint. While the cellular prevalences do seem to correlate roughly to VAF, clinically it doesn't make sense in this patient to have had this ALK mutation at all prior to that sample timepoint.

For this analysis, I ensured that for the union of mutations from all samples, any mutations that are not shared by a particular sample timepoint are set to 0 var_counts and ref_counts corresponding to the corresponding depth in the BAM file at that genomic position. I used tumor purity and copy number estimates from FACETS.

Any help would be greatly appreciated! Thanks!

[nihilee]

@jmchan88 It seems I met the same problem? have you figured out the reason? Any help would be greatly appreciated! Thanks!