error if too many (contaminant) contigs about shiver

[ChrisHIV]

Hi Carlijn, thanks for the interest. That sounds like something I can fix, though I'm afraid probably not before Jan 29th.

[CarlijnB]

Thanks Chris!
A quick google appeared to suggest it was the length of the argument, rather than the number of separate arguments, that would be the issue.

[ChrisHIV]

Hi Carlijn, sorry for the delay. I think the latest pushed change should solve the issue - the contig names are now passed between steps via a file. Let me know...

[CarlijnB]

Hi Chris,
Thanks for looking at this. I've just downloaded shiver-master and have been playing around with it.

Sorry, I don't actually know yet whether your original fix works - I'm now getting different issues.
The same test data is now giving me bother about trimmomatic ("Error running 'trimmomatic -version"), despite having config.sh set to TrimReadsForAdaptersAndQual=false.
My reads should already have been cleaned, and I do not have trimmomatic installed. Any quick fix to tell shiver this, other than just installing trimmomatic regardless?

Another set of data has issues shortly after BLASTing contigs against ref alignments. It generates the .blast file, but then I get:
/home/s1144826/progs/shiver-master/shiver_funcs.sh: line 617: HIVContigsListOrig: unbound variable
/home/s1144826/progs/shiver-master/shiver_funcs.sh: line 618: HIVContigsListOrig: unbound variable
It subsequently just quits/moves on to the next sample without further message.
Any advice?

Thanks again!
Carlijn

[ChrisHIV]

Hi Carlijn, sorry for the delay. I'm guessing what's happening here is that you've set TrimReadsForAdaptersAndQual=false in your new version of config.sh, but you're providing an old version of config.sh as the argument when you run a shiver command. Can you confirm/refute? I can't see from the code how you would get the former error unless TrimReadsForAdaptersAndQual=true.