Clarification on Input Files and Clone Number for runmtclone.py (MityBayes)

[Jacky-Yiu]

Hi,

I am currently working on identifying clonal types in my DLP library using mitochondrial variants, and I would like to use your runmtclone.py script (referred to as MityBayes in your paper, I believe). I have a few questions about the input:

From reviewing the code, it seems that two files, total.csv and allele.csv, are required. Are these matrices where rows represent SNVs and columns represent cell IDs, with the values being the total read depth (DP) and allele depth (AD), respectively?

Regarding the num_of_clones parameter, do I need to provide the exact number of clones I expect, or is this just the maximum number of clones? I’m uncertain about the exact number of clones in my data.

Lastly, is the runmtclone.py script the final version of MityBayes, or is there another repository or version I should refer to for the latest implementation?

Thank you for your help!

[mik2044]

Hi,

Thanks for reaching out! I'll address your questions below:

1. Input files:
   Yes, you're correct. Rows = SNVs, Columns = Cell IDs.
   total.csv: Total read depth (DP) at each variant site.
   allele.csv: Allele depth (AD) for the alternate allele at each variant site.

2. num_of_clones parameter:
   This serves as a prior and is usually the maximum number of clones you expect. It basically shouldn’t exceed the number of SNVs in the data. I recommend testing a range of values (e.g., 2-10) and scoring the results. Check for even clustering or compare with a nuDNA copy number-based method using the Adjusted Rand Index (ARI) or the elbow plot of the mean silhouette score.

3. Script version:
   Yes, the runmtclone.py script is the latest version of MityBayes at this time.

I hope this helps! Feel free to reach out with any further questions.