Running analysis on sample sample id's

[Indianhedgehog]

Would it be possible to run pairwise analysis on same id samples only ?
At present the workflow accepts test1, test2, but, let's say you have same id H021-YAGAH and H021-YAGAH, but output from different callers. would it be possible to run?

[kubranarci]

Hello @Indianhedgehog, the input parameter for id is not actually representing sample id in this case. It is called as "id", suppose to represent unique id per test case. I think keeping it the same as it is important to keep sanity of the workflow. if you have same samples called using different callers, you should, unfortunately, name something like H021-YAGAH_caller1 vs H021-YAGAH_caller2 or H021-YAGAH_1 vs H021-YAGAH_2. This is important since id is being used to tag cases in the plots, if we use some other combinations directly through the pipeline, I believe it will make tags longer than needed at some cases.

[Indianhedgehog]

Hi @kubranarci , if we use the combination you suggested, it does pairwise comparison of all the samples if other sample ids are also present.
For example: H021-YAGAH_caller1_H021-YAGAH_caller2 and H021-YAGAH_caller1_H021-GHDFD_caller1
How do I avoid that?

[kubranarci]

Pairwise comparison is only for "concordance" analysis, not for other methods.