Testing Lirical 2.0, and getting VERY different results from 1.3.4 version

TheJacksonLaboratory / LIRICAL

LIkelihood Ratio Interpretation of Clinical AbnormaLities

https://thejacksonlaboratory.github.io/LIRICAL/stable

Other

22 stars 11 forks source link

Testing Lirical 2.0, and getting VERY different results from 1.3.4 version #650

Closed dovelleiro closed 2 months ago

dovelleiro commented 5 months ago

Dear guys, Lirical is an important part of my pipeline for rare diseases in neuromuscular diseases. With the new version, I'm obtaining VERY different results. In a sample I'm using as reference (the causal variant has been verified in functional analyses), that variant was found as the first option using Lirical 1.3.4 (and other 4 prioritization tools). Using Lirical 2.0, the same variant is demoted to the 29th position. I can give you more details by email. Best, David

ielis commented 5 months ago

Hi @dovelleiro that is unexpected because I compared the benchmarks and got slightly better results for 2.0.0, likely due to incremental updates to HPO resources.

Can you please share more details?

Thanks a lot and all the best..

dovelleiro commented 5 months ago

I've sent you an email to your Proton address. D

ielis commented 4 months ago

Hi, sorry for taking a longer time to respond. I replied to your mail.

ielis commented 2 months ago

Fixed by #651

@dovelleiro please reopen if the issues persist. Thank you!