average quality score for fastq file

[jgibbons1]

Hello,
Simple question here, but I've had a problem finding a program to do it.

Given a fastq file, is there software that will calculate the average quality score for the entire file?

Thanks

[JackieBadger]

The "Compute quality statistics" function in Galaxy (https://main.g2.bx.psu.edu)

[jgibbons1]

Thanks @JackieBadger

Do you know of any unix/linux based tools to do this? I'm trying to stay away from galaxy for this tidbit.

[JackieBadger]

Thanks @JackieBadger

Do you know of any unix/linux based tools to do this? I'm trying to stay away from galaxy for this tidbit.

Why do you need to avoid Galaxy?

[jgibbons1]

I'm trying to integrate this value into a larger analysis pipeline. Automating it would be more efficient.

[GenoMax]

Hello,

Given a fastq file, is there software that will calculate the average quality score for the entire file?

Thanks

One should be careful about "average quality score" for an entire file since you could still have a subset of sequences that may be hidden outliers ("bad") in an otherwise "good" file.

Both Fastx toolkit (http://hannonlab.cshl.edu/fastx_toolkit/) and FastQC (http://www.bioinformatics.babraham.a...ojects/fastqc/) will generate statistics you could use.

If this is illumina data and you have access to CASAVA pipeline output then the summary file has a "mean" quality score for each sample that you could parse.

[JackieBadger]

You can also get the source code for any Galaxy function

[maasha]

With Biopieces (www.biopieces.org) you can do it in two steps where you first find the mean score per sequence entry and then the overall mean score:

read_fastq -i test.fq | mean_scores | mean_vals -k SCORES_MEAN -x

[jgibbons1]

Thanks to everyone's suggestions!

I will try a few of these and let you know what I ended up doing.

[jgibbons1]

For anyone still interested...

For simplicity, I ended up using Bioawk (https://github.com/lh3/bioawk):

Get the mean Phred quality score from FASTQ:
awk -c fastx '{ print ">"$name; print meanqual($qual) }' seq.fq.gz

[xApple]

That bioawk command doesn't compute the quality score for the entire file as per the question you originally asked. It computes it for every sequence and floods your standard output.

Here is a program that actually does what you asked.

#!/usr/bin/env python

"""
Compute the average quality for a given FASTQ file.

Written by Lucas Sinclair.
Kopimi.

You can use this script from the shell like this:
$ fastq_avg_qual < reads.fastq
"""

# Iterative mean #
def imean(numbers):
    count = 0
    total = 0
    for num in numbers:
        count += 1
        total += num
    return float(total)/count

# Do it #
import sys
from Bio import SeqIO
records = (r for r in SeqIO.parse(sys.stdin, "fastq"))
scores = (s for r in records for s in r.letter_annotations["phred_quality"])
print imean(scores)

[jgibbons1]

Thanks xApple. You're right...I ended up taking the average of the quality scores.

I will give your script a shot. Thanks!

[xApple]

You're welcome ! It's just a quick hack that has the advantage to not use any memory but suffers from being extremely slow. Probably not suitable for Illumina sized data.