fastq_quality_filter -p 100 or could be less?

Sergio.pv · 04-02-2014, 06:52 AM

Hello everyone,
I am using FASTX-toolkit to filter my data (small RNA fraction for miRNA analyses) according to quality.
I use the following settings:
fastq_quality_filter -v -q 17 -p 100 -Q 33

This means, that the output will consist in reads where 100% nucleotides are >=17 phred score.

However, when doing that I loss a big amount of reads. For example:
Quality cut-off: 17
Minimum percentage: 100
Input: 95428 reads.
Output: 29635 reads.
discarded 65793 (68%) low-quality reads.

My question is, whether is there a safe trade off between using lower minimum pecentage and still having a good file for mapping.

Thanks in advance

dpryan · 04-02-2014, 07:01 AM

Q17 is a pretty strenuous threshold, you might try 5, or better yet just do regular trimming and discard anything too short to map.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
100 genome phylogeny	mdk308	Bioinformatics	4	09-25-2013 12:32 PM
Fastq_quality_filter no removing sequences	oneofmany	Bioinformatics	2	08-28-2013 05:09 PM
fastx: fastq_quality_filter	daiello	Bioinformatics	4	05-31-2013 11:23 PM
FastX: fastq_quality_filter problem	ElMichael	Bioinformatics	11	11-09-2012 12:19 AM
What does mean 2 x 35 or 2 x 100 in illumina solexa?	Stock	General	1	02-14-2010 03:09 AM

04-02-2014, 06:52 AM	#1
Sergio.pv Member Location: Berlin Join Date: Jul 2013 Posts: 20	fastq_quality_filter -p 100 or could be less? Hello everyone, I am using FASTX-toolkit to filter my data (small RNA fraction for miRNA analyses) according to quality. I use the following settings: fastq_quality_filter -v -q 17 -p 100 -Q 33 This means, that the output will consist in reads where 100% nucleotides are >=17 phred score. However, when doing that I loss a big amount of reads. For example: Quality cut-off: 17 Minimum percentage: 100 Input: 95428 reads. Output: 29635 reads. discarded 65793 (68%) low-quality reads. My question is, whether is there a safe trade off between using lower minimum pecentage and still having a good file for mapping. Thanks in advance

04-02-2014, 07:01 AM	#2
dpryan Devon Ryan Location: Freiburg, Germany Join Date: Jul 2011 Posts: 3,480	Q17 is a pretty strenuous threshold, you might try 5, or better yet just do regular trimming and discard anything too short to map.