You are here

FloppyTail :How to specify range of design(mutation)

7 posts / 0 new
Last post
FloppyTail :How to specify range of design(mutation)
#1

Hello 

 

I am working with Floppy tail.

I am planning to do Floppytail with the amino acid design in specified range (250 to 255 in Rosetta count at chain B protein).

Is that possible to specify the design range by resfile like below? 

I couldn't fully understand packer task control......

 

## inside of resfile##

250 B ALLAA

251 B ALLAA

252 B ALLAA

253 B ALLAA

254 B ALLAA

255 B ALLAA

 

I used Floppy tail with below command and mapmover file

mpiexec -np 8
FloppyTail.mpi.linuxgccrelease
-in:file:s input.pdb
-in:file:movemap mapmover
-FloppyTail::perturb_temp 0.8        
-FloppyTail::perturb_cycles 5000
-FloppyTail::refine_temp 0.8
-FloppyTail::refine_cycles 5000
-FloppyTail::refine_repack_cycles 30
-nstruct 1500
 


##inside of mapmover##
RESIDUE 250 255 BBCHI

 

 

 

Post Situation: 
Mon, 2015-11-02 04:14
syoifczeri

Your resfile would be better if it read:

## inside of resfile##

NATAA

start

250 B ALLAA

251 B ALLAA

252 B ALLAA

253 B ALLAA

254 B ALLAA

255 B ALLAA

 

The extra header allows for repacking outside the tail region.  

 

You should run this NOT in MPI, with maybe 100 cycles, for just a few nstruct first - this will tell you if all the flags work (does it crash?  does the sequence come out different?) without costing much computer time.

 

With -np 8 for mpi, you should use the flag "-jd2:mpi_work_partition_job_distributor true" for more efficiency on few processors.

 

For better packing in your "production run", you probably ought to use -use_input_sc and -ex1 -ex2 for more rotamers.

Mon, 2015-11-02 07:27
smlewis

Hi.

Thank you for your advice.

I tried , but I got error message as below.

ERROR:: Exit from: src/core/pack/task/ResfileReader.cc line: 1508
 

I couldn't find out the cause of read error.

Would you give me more suggestion?

 

####commandtest 

mpiexec -np 8 FloppyTail.mpi.linuxgccrelease

-jd2:mpi_work_partition_job_distributor true

-in:file:s input.pdb

-in:file:movemap mapmover

-FloppyTail::perturb_temp 0.8

-FloppyTail::perturb_cycles 100

-FloppyTail::refine_temp 0.8

-FloppyTail::refine_cycles 100

-FloppyTail::refine_repack_cycles 30

-resfile resfile

-nstruct 8

 

###resfile

NATAA
EX 1 EX 2
USE_INPUT_SC
start
250 B ALLAA
251 B ALLAA
252 B ALLAA
253 B ALLAA
254 B ALLAA
255 B ALLAA

 

###mapmover

RESIDUE 250 255 BBCHI

 

 

 

Wed, 2015-11-04 21:52
syoifczeri

A) If you want to use ex1 ex2 use_input_sc (and you should), put them as command line flags, not in the resfile.  Putting them in the resfile has the curious side effect of NOT using those flags on residues 250-255, because they aren't specified there.  

 

B) Your resfile looks fine at a glance, post more of the error to help me diagnose it.

Thu, 2015-11-05 10:24
smlewis

Hi

thanks for your prompt reply.

I also tried putting -use_input_sc -ex1 -ex2 in command line. but failed.

I thought some 2byte character contaminated in resfile and I made new resfile2. but, same error occured.

 

I have read the strings near onError(err_msg.str()); in ResfileReader.cc but I couldn't find what error aborted the process.

So, I attached the two log files (Sorry for long log files........). 

 

Best regards

 

File attachments: 
Thu, 2015-11-05 19:12
syoifczeri

Your log file is long because you're running in MPI.  Let me suggest that you run on ONE processor, not in MPI, until the bugs are worked out, then scale up to production runs on many processors.  You can also try -mpi_tracer_to_file [filename], which will write each process's log output to its own file.

 

The error: "core.pack.task.ResfileReader: (5) On line 3, the pose does not have residue with chain=B, PDBnum=96."  Is there a chain B, residue 96 in your input PDB?

Fri, 2015-11-06 13:14
smlewis

Hello

 

 

 

> Is there a chain B, residue 96 in your input PDB?

Oh! chain B have no 96 residues!

 I subtracted a number of amino acid length of chain A from number in resfile . Then,,,,Program run without any problem.

 

Thank you for giving me good advice.

 

Sat, 2015-11-07 21:42
syoifczeri