Page 1 of 1

Replicas picked for clustering and model genaration while using I-TASSER

Posted: Tue Jan 10, 2023 10:27 am
by rupesh
Hi, I have a doubt regarding the intricate methodology of I-TASSER standalone program. After performing the MC simulation via an A type run (ab initio), 8 replicas were picked for clustering, which is in contrast to only 3 replicas that were picked after an M type run.
It would be great if someone can explain the reasoning behind these two different approaches and how the replica numbers 8 and 3 were selected for the two different types of simulations.

Re: Replicas picked for clustering and model genaration while using I-TASSER

Posted: Wed Jan 11, 2023 1:06 am
by jlspzw
Dear user,

Thank you for using the I-TASSER. You can find some explanations on our CASP14 abstract, https://predictioncenter.org/casp14/doc ... tracts.pdf. Please see the Zhang-Server section.

The numbers 8 and 3 are decided based on a benchmark test that aims to optimize the final model quality after the SPICKER step.

Best
IT Team

Re: Replicas picked for clustering and model genaration while using I-TASSER

Posted: Thu Jan 12, 2023 6:35 am
by rupesh
Thanks for your prompt response. It is much appreciated!