RosettaCommons - Fragment Generation

Forums:

Rosetta 3 - General

Post Situation:

Memory Leak in FragmentPicker.cxx11thread application

Corvin — Tue, 11 Jul 2023 13:04:04 +0000

Dear Sir and Madam,

I have a question regarding memory leak in FragmentPicker.cxx11thread application.

When I launch Rosetta FragmentPicker thread application on SLURM (with all relevant frags:non-local and frags:contacts settings turned on), the amount of consumed memory grows up fastly, cluster gets overloaded and application's run is aborted (st9bad_alloc) with ExitCode 134.

I noticed, that with bigger number of thread this memory growth usage gets faster, consequently, the software stops even sooner. On this basis I may suppose, that it is thread-related problem. I read, that the problem often consists in some features of C++.

I have tried to restrict the input and amount of processed data by a list of denied_pdb and cutting the number of candidated and selected fragments per position, respectively. I would notice, that the foregoing error (or stucking at this stage) happens at the saving fragments stage.

UPD:

But after thorough examination of slurm[jobid].out file I noticed, that this memory usage increases even after closing all threads, when there is only one thread left (I mean, that application is still running). At least, slurm[jobid].out doesn't present any further multithread activity despite the highest possible log output ('trace') level 500 - probably the progress displaying stucks at some stage. That contradicts, what I see at cluster - htop shows usage of specified number of CPUs until the very end.

Would you be kind to tell me, how to fix this problem, please?

I will be sincerely grateful for your response.

Best regards,

Corvin.

Category:

Forums:

Post Situation:

Generating a new Vall database

vijay.anand — Tue, 20 Jun 2023 19:08:04 +0000

Hi, I've been looking through the fragment-picker application and my objective is to select fragments from the vall database such that only fragments that are 100% identical to the query fragments are selected. Since the probability of getting larger fragments, that are entirely identical to the query fragments, is low; I wish to create a newer updated vall database from more recent and bigger protein databases. Could someone guide me through achieving this task or point me to the required references?

Category:

Forums:

Post Situation:

Ways to improve AbRelax performance?

Corvin — Tue, 19 Apr 2022 13:10:28 +0000

Dear Sir and Madam,

I would like to ask you regarding the ways to improve the performance for either FragmentPicker and AbRelax applications.

I have already done a lot of work, trying to pick the right settings and relevant to them values for both FragmentPicker and AbRelax. But, unfortunately, I haven't succeeded in generation of protein structures.

They either don't pass turned on filters, or AbRelax produces poorly folded structures without proper secondary structure.
When I turn on nonlocal setting for FragmentPicker, it is not able to produce fragment files for feasible time (calculation takes too much time, e.g. there are not still results even after 10-12 hours of work).

In addition, I ran these application through MPI or use FragmentPicker compiled under cxx11thread.

Nevertheless, these approaches haven't led me to successful outcome yet.

I would hazard to guess, that the problem lies in computational capacities of hardware. I must note that my laptop has 8 Intel i5 CPUs with 1.6 GHz performance for each one. I have read such a claim, that Rosetta is not good at modelling ab initio of protein, longer than 150 AAs. My protein consists of 264 AAs.

Would anyone be kind to tell me either the way to improve my option settings or I should just to find some more powerful hardware, please?

Best regards,

Corvin.

Attachment	Size
AbRelax protocol	2.25 KB
FragmentPick protocol	1.27 KB

Category:

Structure prediction

Post Situation:

Forums:

Chemically Modified Residues

Weekly releases for commercial users?

rjacak — Tue, 22 Mar 2022 14:48:27 +0000

Hi Steven, Rocco, et al.,

Would it be possible to have the weekly releases for commercial users restarted? The last weekly release available (for commercial users, maybe commercial and academic users?) is Rosetta 2021.16, with a release date of Friday, April 23, 2021. I know there have been bug fixes and other minor improvements made to apps that I am using and it would be great to have access to those. If weekly releases are unlikely to restart anytime soon, would it be possible to have the source code changes of bug fixes merged into main (master) posted somewhere so that licensees can patch the source themselves?

Thanks,
-Ron

Category:

Forums:

Rosetta 3 - General

Post Situation:

Generating the 3-mer and 9-mer fragments

Prasoon — Wed, 09 Feb 2022 17:04:52 +0000

Hello everyone,

I have recently started using Rosetta and I wanted to generate the 3-mer and 9-mer fragments for Abinitio (without using the Robetta server).
I used the make_fragment.pl files provided under the fragment_tools folder in Rosetta to generate the files (by giving just the fasta file as an input). However, the result that I gathered was not the same, as what is generated by the Robetta server.

I would really appreciate any help on what I am doing wrong or how I can verify the results or are there any other ways to generate the 3-mer, 9-mer file without having to use the Robetta server.

In addition to the configuration changes in the make_fragment.pl file, I once changed the default database from nr to uniref50, but I later reversed it.

Thank you.

Category:

Forums:

Post Situation:

Chemically Modified Residues

RAbD with N-term，Proline-like ncaa returns segmentation fault during CCDEndsGraftMover

CordycepsLin — Mon, 07 Feb 2022 03:09:18 +0000

Recently I'm stumbling with rosetta RAbD application (version 2020.08.61146) in hope to modify an antibody whose antigen containing a non-canonical,N-term, and PRO-like aa ,namely the pyroglutamic acid in its epitope.

I generated params file according tutorials in :https://new.rosettacommons.org/demos/latest/public/design_with_ncaa/README,with input mol file below.

the product params runs properly during dock,relax,and RAbD's seq_design mode (except for wrong number in N-term hydrogen adding). but when doing graft design ,the application failed during CCDEndsGraftMover with a segmentation fault.

I had a series of attempts to solve the problem:

I fixed the patch file in database/chemical/residue_type_sets/fa_standard( and centroid ) /patches .

I tried to use Pro's centroid Params for this one. I tried to modify the ICOOR_INTERNAL lines so as to make the fold tree grows from CA (rather than from N , as the python script generated) ,adding virtual N and cut_bonds(as Pro's params does) .

I tried to change the value of NBR_RADIUS.

I even tried to modifying the params directly from PRO's by adding an acyl-oxgen atoms.

Below are two files I finally got. Sadly neither of them works. Due to limit error message, it's quite hard for me to debug the 'segmentation fault'.

Can anyone give it a look and help me out ?

thank you very much.

Attachment	Size
params_modified_from_PRO.txt	2.36 KB
params_gen_and_modified_manually.txt	1.99 KB
starting_pose_for_design.txt	145.47 KB

Category:

Design

Post Situation:

Forums:

Robetta

Rodriguez — Fri, 25 Jun 2021 18:34:50 +0000

I want to create Robetta locally on my high-powered home computer. Does anyone have detailed instructions on how to do this?

I would appreciate any help!

Category:

Rosetta 3 - Build/Install

Forums:

Post Situation:

r_frag_quality

ii_cnr — Mon, 12 Apr 2021 10:06:11 +0000

Hi everyone,

I would like to assess my designs using the fragment quality metric and biased forward folding. In both cases Rosetta documentation suggests to use the r_frag_quality application. However I cannot find it in my locally compiled versions of rosetta (for example in rosetta_src_2020.08.61146_bundle). Can anybody help?

Thanks

Category:

Forums:

Post Situation:

Solved

Creating fragment files for proteins less than 27 aa in length

Poladu — Tue, 26 Jan 2021 15:50:52 +0000

Good day.
I'd like to make abinitio structure prediction using rosetta of some small proteins (e.g. 2jof, which length is 20 amino acids).
Such prediction requires fragment files, which can be generated using web Robetta Server or by youself using PSI-BLAST.
However, when i tried to submit a job on Robetta there was an error:
Sequence length must be between 27 and 1000 residues
Similar restriction was described in rosetta manual dedicated fragment-files:
Obtain a fasta file for the desired sequence. This file must have 60 characters/line with no white space. Therefore, no less than 60 aa.
So, i'd like to ask if there is some way to generate fragment files for proteins, which length less than 27, in order to predict their structure using Rosetta Abinitio?

Many thanks in advance.

Category:

Forums:

Post Situation: