xhhuanglab / OutcrossSeq

11 stars 10 forks source link

Heterozygosity rate #2

Open Yutang-ETH opened 3 years ago

Yutang-ETH commented 3 years ago

Hi,

First of all, thank you very much for developing these nice scripts. I just have a stupid question regarding the heterozygosity rate. How can I calculate it? Thank you very much.

Best wishes, Yutang

mjchen-st commented 3 years ago

Hi, I don't understand what you mean very much. If you simply calculate the heterozygosity rate, you can divide the heterozygous loci in vcf by the total number of loci.

Yutang-ETH commented 3 years ago

Hi,

Thank you very much. OK, I see. Actually, my question is in your example command line, how do you get the 0.3?

Best wishes, Yutang

mjchen-st commented 3 years ago

In the algorithm processing, the highly heterozygous sites are equivalent to noise in the calculation and should be removed. I am considering the number of markers that can be retained for subsequent calculations in each segment to determine the heterozygosity rate. For example, there are 1000 per 100kb.

2021-09-16

15021029575

发件人:Yutang-ETH @.> 发送时间:2021-09-16 20:38 主题:Re: [xhhuanglab/OutcrossSeq] Heterozygosity rate (#2) @.> @.**@.>

Hi, Thank you very much. OK, I see. Actually, my question is in your example command line, how do you get the 0.3? Best wishes, Yutang — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.