You are here

Rotamers used in fixbb

7 posts / 0 new
Last post
Rotamers used in fixbb
#1

Dear friends,
I am trying to use fixbb to replace some amino acid residues from a template PDB into my own sequence.

Can I ask
1) Is "-dun10" the latest rotamer database? (How can I know if it is the latest?)
2) Do I need to add additional flags (e.g. -ex1, -ex2, -ex3, -ex4) or they have been set by default?

Basically, I do not have particular requirement for the structure. So I am wondering if all the flags have been set by default?

The command line is:
~/Cheng/rosetta_2014.30.57114_bundle/main/source/bin/fixbb.linuxgccrelease @/home/lanselibai/Cheng/fixbb/flags_fullatom_dun10 -database ~/Cheng/rosetta_2014.30.57114_bundle/main/database -resfile /home/lanselibai/Cheng/fixbb/LC_resfile.txt >& /home/lanselibai/Cheng/fixbb/log.txt &

The options file is:
-s /home/lanselibai/Cheng/fixbb/4KMT_LC_native_442residue_FullAtom.pdb
-dun10
# I jus want one output structure as I will do relax later.

Thank you very much.

Yours sincerely
Cheng

Post Situation: 
Thu, 2014-11-06 02:47
lanselibai

The default rotamer library is the latest.

Common packing flags include: -ex1 -ex2 -extrachi_cutoff 0 -use_input_sc. The first three test off-rotamer rotamers, and the last allows the input structure as a rotamer.

If you really want just one structure to relax, then what you have is fine.

Fri, 2014-11-07 12:06
smlewis

Hi smlewis,
Thank you. If I understand correctly,

1) I do not need to add "-dun10" as the default is already the latest rotamer database.
2) I also do not need to include "-ex1 -ex2 -extrachi_cutoff 0 -use_input_sc" if I just want one output.
3) It would be better to include "-ex1 -ex2 -extrachi_cutoff 0 -use_input_sc" if I want to have multiple outputs.

Are these correct?

Thank you.

Yours sincerely
Cheng

Fri, 2014-11-07 13:39
lanselibai

I think you do not need to include "-ex1 -ex2 -extrachi_cutoff 0 -use_input_sc" because you will do relax afterwards, which does exactly the same thing, if not more thorough, NOT because you only need one decoy. Similarly, if you generate multiple decoys, you do not need to include those options if you do relax later. (But I personally do not see necessity to generate more decoys if the mutations are unique.)

Tue, 2014-11-11 03:02
attesor

Hi attesor,
Thank you. So I will not include those options and only generate one decoy followed by relax.

Yours sincerely
Cheng

Tue, 2014-11-11 06:11
lanselibai

The current defaults are to use the 2010 Dunbrack library. (So "-dun10" defaults to being on. You do not need to add it, but it will not hurt if you do.)

Tue, 2014-11-11 14:22
rmoretti

Hi R Moretti,
Thanks a lot. That would be cool. So in my simple case, resfile is the only file that should be prepared except the input PDB file.

Yours sincerely
Cheng

Tue, 2014-11-11 15:28
lanselibai