The sum of prop is not equal to 1 in subclon.csv

[DayTimeMouse]

Hi,

I used the pyclone format tsv file as the fastclone's input. The output subclone.csv，sum of prop is not equal to 1 （subclone : prop, 0: 0.286945820154048; 1: 0.858586430702103）, is this mean there are two subclones?

But because of the heterogeneity of the cancer sample, it is not reasonable to infer that there are only two subclones.

Best wishes.

[Yaoyx]

Hi @DayTimeMouse,

Thank you for using FastClone!

To the first question: Yes, it shows two subclones. FastClone tries to create a Phylogeny tree, so one subclone could be a parent subclone. (subclone 0 might develop from subclone 1, and this depends on if they share enough mutations. For details, you can check FastClone paper). Hence, the sum of prop is not equal to 1.

To the second question: it could be more than two subclones. However, I think it really depends on the threshold. Let's put it into a simple example, if we assume as long as two cells have one different mutation should belong to two subclones, then we can get as many subclones as possible. Go back to FastClone, its inference of subclone number is based on the number of peaks found from the distribution of cell proportion, so from the distribution the algorithm only captures two obvious peaks. I think there might be better threshold for subclone detection from latest literature, and I would love to know!

Regards
Yao