search

hall-lab / speedseq

A flexible framework for rapid genome analysis and interpretation
MIT License
311 stars 116 forks source link

how many samples are needed for stable variant calling? #112

Open ieiwk opened 7 years ago

ieiwk commented 7 years ago

Dear cc,

I found that the variant calling results of a perticular sample is different when its bam file is used seperately or together with others, and I found results called together with others' bams are more reliable.

Now my question is, how many samples are needed for stable variant calling? If there are new data, do I need to process them together with all processed data together?

Thanks a lot!