Should I do any quality filtering in such a case?

[GloriaFu]

Hello!

I'm sooooo new in NGS field (only 2-day-experience) and having a problem in judging data quality. In this case (see the figure), should I do any quality filtering or trimming? The first quartile seems very low, even to 8. Is this data good enough to use directly?


Thank you for helping so much!

the figure is here: http://www.wretch.cc/album/show.php?...323235.jpg&p=0

[DZhang]

Hi GloriaFu,

What's your intended application? Re-sequencing? de novo assembly? The quality of your data is OK but not good or great. I would definitely perform quality trimming or filtering.

Douglas

https://www.contigexpress.com

[GloriaFu]

Hello

Thanks for reply. These reads are from transcriptome and will be used in expressional analysis. I'm thinking to trim the 10-15 last bps. Does that make sense? Do I have alternative solution in quality filtering?

Thanks so much!

[DZhang]

Hi GloriaFu,

You have other solutions and there are quite a few out there. My usual picks are SolexQA (it has perl scripts for quality-based trimming) and FASTX toolkits.

Douglas

https://www.contigexpress.com

[volks]

quality isnt so bad. a value of 25 is still quite high.
if you insist on doing quality trimming you could use cutadapt which has BWA-like trimming option (if you dont use BWA for alignment ..).