Fastx Toolkit for Quality Stats of data from new illumina pipeline software

[vedbar]

Hi,

My data is in a new format generated by the new illumina pipeline software v 1.8. So, the quality scores are in Sanger FASTQ format with ASCII offset 33.

I used following code of Fastx tooklit to find quality stats of my data.

root@SNP215-01:Cold_Shock_Filtered# fastx_quality_stats -Q33 -i Cold_Shock_negative_2.fq -o Cold_Shock_negative_2.fq.stats

I received the following ERROR message:

fastx_quality_stats: Invalid quality score value (char 'J' ord 74 quality value 41) on line 20

So, please help me to resolve this problem.

Thank you for your time !

Vedbar

[maubp]

Illumina also improved the chemistry, so the PHRED score of 41 is real but higher that people are used to, 0 - 40.

I thought fastx fixed this - which version are you using?

[vedbar]

I am using FASTX toolkit 0.0.12.

[swbarnes2]

Something like:

sed s/J/I/ oldfastq.fq > newfastq.fq

Will change all the 'J's to 'I's; that's a quick and dirty workaround. You'll be slightly lying about how awesomely high quality your letters are, because J's are a little better than I's, but it shouldn't make much difference.

[maubp]

I am using FASTX toolkit 0.0.12.

According to http://hannonlab.cshl.edu/fastx_toolkit/ the latest release is FASTX 0.0.13, but that doesn't explicitly mention this issue. Given the release date (Feb 2010) then I doubt it will cope.

So try v0.0.13 and if it also fails please contact the FASTX authors.

[kmcarr]

According to http://hannonlab.cshl.edu/fastx_toolkit/ the latest release is FASTX 0.0.13, but that doesn't explicitly mention this issue. Given the release date (Feb 2010) then I doubt it will cope.

So try v0.0.13 and if it also fails please contact the FASTX authors.

Yes, v0.0.13 does fix this problem. See this thread/post for a more detailed explanation:

http://seqanswers.com/forums/showthread.php?p=49667

[vedbar]

Thank you !