This is a test version of Biostars. For the public version, visit https://www.biostars.org.
GC-BIAS correction for BAM file

Hi,

I am looking for any software which can correct the GC-Bias in the BAM file and provides the output as BAM.

I know that deeptools (correctGCBias) does the exactly same job, but the output of the deeptools is variable in each run.

So, I am looking for more specific tool.

Basically, I want to count the sequence reads for each chromosome in NIPT (low depth, WGS sample) BAM file after the GC-Bias correction.

Many LOESS based or R-based GC-bias correction methods are available, but they provides the output as normalized data, not as total number of reads in each chromosome/ specified region.

Any suggestions?

genome rna-seq r deeptools
• 5,273 views • link
updated by Biostar Biostar • written by amitgsir amitgsir

1 answer

The reason correctGCBias can produce slightly different results each time is due to using a random number at one point. If you really need to produce the exact same results each time then you could hardcode a seed in. I expect that adding the following near the top of deeptools/correctGCBias.py will do that:

np.random.seed(1234)
• 0 views • link
written by Devon Ryan Devon Ryan

Thanks Ryan, I will try this and will inform you if this works.

• 0 views • link
written by amitgsir amitgsir

Yeah,

Fixing the seed using np.random.seed(1234) is giving the exact same results each time.

Thanks

• 0 views • link
written by amitgsir amitgsir

Great, glad that worked!

• 0 views • link
written by Devon Ryan Devon Ryan

@ Devon This post of yours and the deeptools paper are both from 2016.

Fast forward to the present day (Jun 2020), and the advice I seek from you is whether correctGCBias from deeptools is still good enough a tool for RNA-Seq GC bias detection and correction?

If not, which software / tool would you recommend? ALPINE or something else?

Thanks, in advance.

• 0 views • link
written by Anand Rao Anand Rao

Log in to answer this question.

More posts like this