No call for exomes analysis

Question

[Hällyss]

Hello,

I'm trying to run CNVkit with 20 exomes of patient with Alzheimer disease. This exomes were screening by cytoscan and have many CNV. I have no négative pool so I make a baseline with this 20 exomes. I use the algorithm Flasso because CBS is not installed in my computer. After I use the recommanded parameters.

And I have no call. (cn != 2)

Have you an idea to have a calling ?

Thank you for your help.

Alice

[WouterDeCoster]

This exomes were screening by cytoscan and have many CNV

How sure are you these are true positives?

[Hällyss]

I would like to believe that all the calls are not all true positive but from there to believe that they are all false positives ....

Answer 1

[Eric T.]

You could try loosening the FDR cutoff with segment -t 0.05 (the default is 0.005). Or try segment -m haar. If most of the CNVs are small, e.g., then I wouldn't expect them to be picked up by segmenting the exome.

[Hällyss]

It appears that the -t 0.05 option will show CNVs. I will continue with my analysis and keep you informed of the results. Thank you

Answer 2

[Eric T.]

As an alternative, I've seen good benchmark results on germline exome data from CLAMMS:

• Source code: https://github.com/rgcgithub/clamms
• Paper: https://academic.oup.com/bioinformatics/article-lookup/doi/10.1093/bioinformatics/btv547

[WouterDeCoster]

Yes, I thought the same and wanted to give it a try a while ago. But the code hasn't been updated since publication and issues on github don't get a reaction/solution.