Question: How to detect GC content bias and correct them
1
gravatar for statfa
20 months ago by
statfa500
statfa500 wrote:

Hi guys,

I was hoping here that if one sample is sequenced in each lane, I don't need to correct gc content bias. But now that I don't know how to find number of samples in each lane, I wanna check if GC content bias has occurred. Should I use deep tools (computeGCBias) on galaxy for my rna-seq bam files and then deep tools (correctGCBias) to correct the bias? There are some statistical methods to correct GC bias in table of read counts too. Which one is better? using deep tools on bam files or using other methods available for counts table? I want to do differential expression analysis.

Thank you

ADD COMMENTlink modified 20 months ago • written 20 months ago by statfa500

Multiple questions with similar question content (Number of samples in each lane in GEO datasets ) is not always fruitful.

If you feel strongly about looking for and correcting GC bias then by all means go for it. You may have to try multiple methods to see what works best for the data you have.

ADD REPLYlink modified 20 months ago • written 20 months ago by genomax73k

It's a different question. There I asked how to find the number of samples in each lane. Nobody knew that. @Devon said if I feel a need to check the bias, then I can do that but he didn't say how. Now I'm asking how. What tools? I searched not only on biostars but on google too. What I learned was to use deep tools. But I still have those questions that I have asked above.

ADD REPLYlink written 20 months ago by statfa500

Fair enough. I assume you have already used CQN (which was what you had used in past thread). Now you have found a couple of tools from deepTools and are following-up.

ADD REPLYlink modified 20 months ago • written 20 months ago by genomax73k

Well, actually, @Devon told me to normalize the data by CQN package. But in fact, we should make sure, first, that there is any bias. I wanna know if I'm right about deep tools detecting the GC-content bias. Then, if I find any bias, removing GC-content bias in bam files using deep tools is better or normalizing the bias in counts table using methods like CQN package?

ADD REPLYlink written 20 months ago by statfa500

I am not familiar with CQN. Perhaps deepTools is using CQN like method under cover to do the correction. I will tag @Devon on this post so he can comment.

ADD REPLYlink modified 20 months ago • written 20 months ago by genomax73k

Thank you very much genomax. I'll appreciate it. Sorry if I bother a lot.

ADD REPLYlink written 20 months ago by statfa500

No bother at all. I learn new things here everyday.

ADD REPLYlink written 20 months ago by genomax73k

And I'm not dumb ;-) I know my posts are available even if I change my username. I changed it because of some other reasons. Don't be so pessimistic. If you look at each of my threads, You will see there are big differences in them ;-)

ADD REPLYlink written 20 months ago by statfa500

Tagging: Devon Ryan

ADD REPLYlink written 20 months ago by genomax73k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 2013 users visited in the last hour