search

humanlongevity / HLA

xHLA: Fast and accurate HLA typing from short read sequence data
Other
101 stars 52 forks source link

An irreproducibility of some results #54

Open PKazakova opened 4 years ago

PKazakova commented 4 years ago

Hello,

I decided to check an reproducibility of results HLA-typing by xHLA. I run xHLA on one sample 10 times and , unfortunately, I got different results. I can suggest that this is due to some difficult regions for alignment (e.g. HLA-B*44:138Q: Evidence for a Confined Deletion and Recombination Events in an Otherwise Unaffected HLA-haplotype). But I can't understand which result is correct and which result isn't. For illustration this case I attach output files for 2 samples with 2 different results. Can you explain why I get different results?

Thanks, Polina. Screenshot_1 Screenshot_2

serge2016 commented 3 years ago

This is important to me too. Is it possible to fix it?