Effective genome size of UCSC hg38

Question

[izzy.yichao.cai]

Hi there!

I am running macs2 for peak calling of ChIP-seq data, but I don't know the effective genome size of UCSC hg38. Is anyone got a hint? The default value of macs2 is for hg18.

Thanks!

[endrebak]

I'd just use the hg18 value though. I made my script for corn genomes etc, where the EGS is not known.

Answer 1

[Bontus]

Have a look at this page for some stats of different genomes

Spoiler for hg38

• total size - 3,209,286,105
• non-N bases - 3,049,315,783 (= mappable / effective size)

Answer 2

[endrebak]

Script to compute the effective genome size: epic-effective

Note that I get slightly different results than some papers, but no one on biostars seem to think my method for computing the egs is wrong.

[izzy.yichao.cai]

I tried to run this. And I came into problems that the code seems not recognize the hg38 fasta genome file from UCSC, because the chromosome name contain "_".

Can you kindly run it for me? My readlength is 101 and I used the reference human genome hg38 from UCSC.

[endrebak]

But you probably do not want to include those chromosomes since they will affect the computation. I would have the same problem. You should remove those "bad" chromosomes, I will try to make the script do it automatically, but do not have time now.