PyClone Copy Number for normal sample

[Yana Vassileva]

Dear all,

I am new to PyClone so I hope the questions do not overlap too much with the previous questions.

My first question is following:

I am evaluating an experiment, where we want to find out what mutations a treatment is causing to an organoid (cell culture) . In this case the original tumour  (OT) becomes  the 'normal' (reference) and the organoid/ cell culture (CC) is the 'tumour'. I am taking the copy number of the OT for the input tsv file in PyClone for the column "normal_cn". However the copy number is not always 2. Do you think that is a problem for the analysis?

The second thing is:

I have some problems with the interpretation of the results. I want to determine how heterogeneous my sample is and to extract the cell clone from the OT, which was selected through the treatment. Is that possible?

I have following plots for the CC:

I am not really sure what to conclude from the plots, except that from the mutations 9 are in once cluster?

I would be very glad about any help!

Best regards

Yana Vassileva

Reserach Group 'Biomedical Genomics'

[Andrew]

Hi Yana,

1. It is not a problem if the normal copy number is not always 2. We commonly use values of 1 for the X and Y chromosomes in males. But other values would be reasonable as well for your experiments.

2. The results looks somewhat strange. There is not a lot of clustering. One suggestion would be to use the beta-binomial prior. It might also be worth trying pyclone-vi (https://github.com/Roth-Lab/pyclone-vi). It's the same model with a slightly faster inference procedure. If the results look drastically different that would suggest the model fitting is struggling.

Is this multiple sample data or just a single sample? To answer your question I would imagine you would want untreated and treated samples. You could then look at a scatter plot of the cellular prevalences of clusters for each time point to see which clusters are changing. I think this might be the last scatter plot you shared? If so it would suggest there is no shift in the clonal composition between samples.

Best wishes,

Andy