Phasing small sample sizes

[sihaybo]

Hi,

I am trying to phase 21 individuals with shapeit5 but I get the error "Less than 50 samples is not enough to get reliable phasing, consider using a reference panel to increase sample size". I was able to phase the same file with shapeit4, but now it is not working. I am using the static version of 5.1.1.

There is no reference panel for the organisms I am working on and sadly I also do not count with more samples. Do I still need to try to create a reference panel or is there a way around this?

Thank you very much in advance,

Sofia

[srubinacci]

Hi Sofia.
For just 21 individuals, you can just go ahead with SHAPEIT4. You won't need to use the advanced functionalities of SHAPEIT5 to handle large amount of samples. SHAPEIT5 phase_common is basically the same algorithm as SHAPEIT4.

Best,

Simone

[sihaybo]

Hi Simone,
Thank you so much for your reply, it is very helpful. I moved to a new cluster and I have not been able to install SHAPEIT4 via conda or compile it. Is there a way to download the SHAPEIT4 static file? I did not find it in the github page.

Thank you again.

Best,

Sofia

[cjjinfinity]

can you suggest a standard protocol for this in shapeit5.1