I am using the standalone version of Rosetta, and so do for the Fragment picker application. in the tutorial, 3 SS prediction programs are used with weightage of 0.6 for psipred and 0.2 each for other two programs; however, the Robetta server, produce results based on only 02 programs, i.e. psipred an Jufo so in this case, what is the weightage for both programs?
please reply to the question.
I'm trying to build a peptide (40 aa) using fragments developed by Robetta as well as using "fragment_picker" application (in bin folder; and following Gront et al., paper ). the results are different (superior) if I use the fragments developed by Robetta webserver as compared locally generated fragments (using standalone version). there is a great difference in the size of the fragment out files (Robetta generated files are bigger).
Could you please tell me specifications of fragment generation procedure used for Robetta web server as I asked in my previous question., i.e. what are the specific values and weight used for Psipred and Jufo predictions.
anyhow I'm using following values:
1 psipred 0.7
2 jufo 0.3
SecondarySimilarity 350 0.5 - psipred
SecondarySimilarity 300 0.5 - jufo
RamaScore 150 1.0 - psipred
RamaScore 150 1.0 - jufo
flags used for fragment generation:
# Input databases
# Query-related input files
-frags::ss_pred /home/gnss/singhma/rosetta/pep2/pep2.psipred.ss2 psipred /home/gnss/singhma/rosetta/pep2/pep2.jufo.ss2 jufo
# Weights file
# What should we do?
# three-mers only, please
-frags::frag_sizes 3 4 5 9
# Quota.def file defines the shares between difefrent quota pools. The total should be 1.0
PS: Here i have one another question, how do Robetta define homologs of teh query sequence (I mean its only identity based or combination of identity and query coverage?)
The last question, The Robetta server has a limitation of query length, I have some peptides lengths 17-20 amino acids, is there any way to use Robetta server or standalone is the lone solution?
Many thanks in advance!
Kindly reply for my queries.