MutSig coverage table

Question

[sadiq.saleem.09]

Hi everyone,

I am trying to use MutSig and would like to know if there is a way/script to generate coverage tables for the experiment under analysis. I have tried to use the standard exome_full192.coverage.txt file provided by Broad Institute but I run into the following error:

silent and nonsilent rates are too different

I guess I will have to create my own coverage tables. I have searched various forums thoroughly including Biostars but so far I haven't been able to find a solution. I will very much appreciate if the folks on this forum, who have some experience with MutSig, can help me out.

Thank you so much!

Sadiq

[ShirleyDai]

Hi Sadiq, Have you tried mutect to generate such a coverage file? BTW, I am searching the same answers. Would you share your solutions if you have one? Thanks Shirley

[April]

Hi Sadiq I came across the same error. I am thinking that may because I used the non-synonymous mutations only. because according to the "mutation_type_dictionary_file", only silent and synonymous mutations are classified as "silent". thus in my data, there are no silent mutations. right now i'm trying to add the synonymous data back, but keep using the exome_full192.coverage.txt what do you think? do you have any suggestions? thank you!

April

update: i did what i mentioned above: adding the synonymous data back. then i ran MutSig successfully and got the data i want

[sadiq.saleem.09]

Hi Shirley and April - Sorry I have not used this tool for many years now. I think I am not the best person to answer this question. Apologies. Wish you all the best

[April]

No problem. I added the synonymous mutations back and it worked. So make sure the input file includes both synonymous and nonsynonymous mutations.

Answer 1

[achristofferson]

CovGen can help make a capture/target specific coverage file. However, your error is most likely caused by how you filter your variants before going into MutSig.

[220173404]

Could you please tell me how to use CovGen?The documentation on the GitHub is too brief and I can't understand.I will be very grateful！