Quality score graph

vruotti · 10-22-2008, 03:26 PM

Just wondering if anyone have seen the same problem with cycle 30.
Thanks,

Victor

new300 · 10-23-2008, 01:21 AM

it's not normal. This is for a single run? If you have the images look at the images, see if they get dimmer or something else strange happens at cycle 30. If you don't have the images then look at the intensity plots, do you have a significant drop in average intensity around cycle 30? What do the error rates for cycle 30+ look like on the control lane?

glacerda · 11-06-2008, 09:25 PM

This is Solexa, right?
If it is, maybe you would like to look at that
maybe you would like to look at that:

Substantial biases in ultra-short read data sets from high-throughput DNA sequencing. 2008, July. Nucl. Acids Res.

[]s
glacerda

TylerBackman · 11-20-2008, 02:06 PM

I already posted this answer in the other thread where you asked about this, but I thought I'd post it here as well. This is possibly an artifact of the Illumina pipeline quality score normalization rather than an actual sudden drop in quality.

This is most likely because the qualities you are looking at are alignment normalized, and a large number of your sequences failed to align to the reference genome (due to a ligated adapter, etc.)

Take a look at the un-normalized scores (s_<lane>_qraw.txt) instead, I think you'll find that the curve is more continuous between cycles.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
454 quality score, z-score,..	nii	454 Pyrosequencing	4	10-15-2020 07:29 AM
Threshold quality score to determine the quality read of ILLUMINA reads problem	edge	Illumina/Solexa	35	11-02-2015 11:31 AM
Threshold quality score to determine the quality read of ILLUMINA reads problem	edge	General	1	09-13-2010 03:22 PM
Different MAPPING QUALITY/PER-BASE QUALITY SCORE	m_elena_bioinfo	Bioinformatics	2	09-02-2010 10:00 AM
Fastq quliaty score and MAQ output quality score	baohua100	Bioinformatics	1	02-19-2009 10:21 AM

10-22-2008, 03:26 PM	#1
vruotti Member Location: US Join Date: Feb 2008 Posts: 13	Quality score graph Just wondering if anyone have seen the same problem with cycle 30. Thanks, Victor

10-23-2008, 01:21 AM	#2
new300 Member Location: northern hemisphere Join Date: Mar 2008 Posts: 50	it's not normal. This is for a single run? If you have the images look at the images, see if they get dimmer or something else strange happens at cycle 30. If you don't have the images then look at the intensity plots, do you have a significant drop in average intensity around cycle 30? What do the error rates for cycle 30+ look like on the control lane?

11-06-2008, 09:25 PM	#3
glacerda Member Location: Brazil Join Date: Aug 2008 Posts: 27	This is Solexa, right? If it is, maybe you would like to look at that maybe you would like to look at that: Substantial biases in ultra-short read data sets from high-throughput DNA sequencing. 2008, July. Nucl. Acids Res. []s glacerda

11-20-2008, 02:06 PM	#4
TylerBackman Member Location: Riverside, CA Join Date: Oct 2008 Posts: 13	I already posted this answer in the other thread where you asked about this, but I thought I'd post it here as well. This is possibly an artifact of the Illumina pipeline quality score normalization rather than an actual sudden drop in quality. This is most likely because the qualities you are looking at are alignment normalized, and a large number of your sequences failed to align to the reference genome (due to a ligated adapter, etc.) Take a look at the un-normalized scores (s_<lane>_qraw.txt) instead, I think you'll find that the curve is more continuous between cycles.