'Use Of Uninitialized Value' Message Using Solexaqa

Question

[Marina Manrique]

Hi all,

I'm running solexaqa in a set of 51 millions of 1x100 Illumina reads, here you are the command I've used

/usr/bin/solexaqa -v -m reads.fastq

the process is working for some hours now and I'm getting this message all the time, I don't know if there's something wrong in the reads file

'Use of uninitialized value in numeric lt (<) at /usr/bin/solexaqa line 546, <INPUT> line 428486886.'

Any of you got this message too? What is your experience with this tool?

Thanks and regards

Marina

[Michael]

Hi, it's hard to speculate about the cause but, this program (solexaq) is seemingly a perl program. This message is a warning not and error, is caused by some code like this undef < 33. So it would be interesting to see what actually is the code in solexaqa at line 546.

It might have to do with the quality values or your fastq file, is is maybe truncated or missing quality values.

[Istvan Albert]

Check the file with FastQC and see if it works with it: http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc/

[Marina Manrique]

@Istvan Albert: Thanks! I'll try FastQC, anyway (as I commented in Michael Dondrup answer) it seems there's a kind of bug in the tool and doesn't support newer Illumina file formats very well...

Answer 1

[Michael]

This program relies completely on fastq records consisting of exactly 4 lines, therefore it is most likely the case that your fastq file contains extra linebreaks and that might drive the program counters out of sync. Can you check if that is the case?

This FASTQ-related problem has been discussed here previously, btw.

[Marina Manrique]

The fastaq format was correct. I asked in the solexaqa google group and got this answer from Murray Cox 'The variance and min/max flags are not very helpful in practice, they take longer to run, and we've been having problems using them in combination with some of the newer Illumina file formats.'