Maq and export format

[Apollo324]

Maq is now being distributed with a tool to convert export format files to standard fastq, and I'm wondering if anyone knows how it treats the reads which failed the Illumina pipeline quality threshold test.

Thanks.

[myrna]

export2std for chastity filtered reads

This function is implemented in the script fq_all2std.pl. I had assumed it would use all reads but later found out that it only took the read (pairs) that passed chastity filtering (there is a Y/N flag in the export file that marks this). You can comment out the line in fq_all2std.pl if you want to include the reads that failed this check:

#if ($t[21] eq 'Y') {
formatting code....
print "+\n$qual\n";
#}

Ryan

[myrna]

export2maq

In a similar vein, I have been hoping to hear from the Maq community whether anyone has had any success running the export2maq function. This is meant to take an export file and reformat it into a .map file (forgoing the need to run maq map or maq assemble). I always get a segmentation fault when I try to use this function (have tried on various platforms). I have gotten no response from the maq-help mailing list, so I thought I would try here.

Thanks,
Ryan

[bioinfosm]

<bump>

I get the same segmentation fault. I had another query, do I need to feed in both the export files for the mate pair (I would expect so).. regardless its a seg fault either ways

[myrna]

I'm glad to see that someone out there is seeing the same issue. I have isolated the problem and made a simple workaround. The converter is expecting certain fields in the export file to be filled, but our pipeline leaves them empty. I pass the exports through a perl script that puts dummy fields in those columns before passing them to export2maq. Note, I pass in both the forward and reverse export files (e.g. cat export_*.txt | add_dummy_values.pl | maq export2maq out.map in.list -)

[bioinfosm]

I'm glad to see that someone out there is seeing the same issue. I have isolated the problem and made a simple workaround. The converter is expecting certain fields in the export file to be filled, but our pipeline leaves them empty. I pass the exports through a perl script that puts dummy fields in those columns before passing them to export2maq. Note, I pass in both the forward and reverse export files (e.g. cat export_*.txt | add_dummy_values.pl | maq export2maq out.map in.list -)

Thanks Ryan.
I had also figured of different formatting for my export output, because of the way my reference sequence for eland was formatted.

Anyways, I guess you are doing downstream analysis to detect structural variants, could you share how you proceed?

I wish to use the maq sv script, but am not getting across the usage..

[lh3]

To bioinfosm:

The input should be something like:

maq mapview -b aln.map | awk '$6!=18&&$9>=10&&$6!=64&&$6!=192&&$6!=130' > sv.inp

You can add more filters in the awk command to control the reads passed through maq.pl sv

[bioinfosm]

To bioinfosm:

The input should be something like:

maq mapview -b aln.map | awk '$6!=18&&$9>=10&&$6!=64&&$6!=192&&$6!=130' > sv.inp

You can add more filters in the awk command to control the reads passed through maq.pl sv

Hi Heng,
I get nothing when using the above filters. Starangly, my 9th column (Quality ?) is 0! Any suggestion as to what might be going wrong there?

and btw, happy new year!

[lh3]

The 9th column is the smaller single-end mapping quality of the two reads in a pair. It is actually the mapping quality of an abnormal pair. You can reconstruct this column anyway.