RosettaCommons - Membrane

Post Situation:

Forums:

Rosetta 3 - General

Docking antibody to membrane protein

Yegor_Ko — Thu, 12 Oct 2023 14:37:53 +0000

Hello, what are the general guidelines for docking, when one of the protieins is a transmembrane and its partner is supposed to dock to the extracellular part (e.g. antibody or ligand)?

I suppose that the transmembrane one should be relaxed using MPFastRelax, whereas the extracellular using the standard REF15?
Which scoring function should be specified for docking?
Should one include any constraint to hide the intracellular and membrane part from docking or it's not necesary?
Are there any example protocols doing that?

Thank you in advance!

Category:

Forums:

Post Situation:

unable to generate the span file with "mp_span_from_pdb.default.linuxgccrelease" from a PDB file

Sajjad — Wed, 30 Nov 2022 19:40:55 +0000

Dear Users,

I am trying to use the "mp_span_from_pdb.default.linuxgccrelease" binary for generating a "span" file to be used as the input for the "predic_ddg.py" python.

However, I face the following error when I start the command.

I face no issues when trying with the demo pdb file. Yet, when I use my pdb (attached), it crashes with the following error.

Could you please help me know what the problem is and how to resolve it?

Thank you

The error:

mp_span_from_pdb.linuxgccrelease -in:file:s mp_ddg/inputs/MC4R_Inactive_model_withNA_A.pdb
core.init: Checking for fconfig files in pwd and ./rosetta/flags
core.init: Rosetta version: rosetta.binary.linux.release-280 r280 2021.16+release.8ee4f02 8ee4f02ac5768a8a339ffada74cb0ff5f778b3e6 https://www.rosettacommons.org 2021-04-20T20:52:25.363712
core.init: command: /media/sajad/18a2b4cf-06c3-437d-bf57-7d68790182b9/sajad/Desktop/Tools/Lab/Modeling/Rosetta/rosetta_bin_linux_3.13_bundle/rosetta_bin_linux_2021.16.61629_bundle/main/source/bin/mp_span_from_pdb.linuxgccrelease -ignore_unrecognized_res -in:file:s mp_ddg/inputs/MC4R_Inactive_model_withNA_A.pdb
basic.random.init_random_generator: 'RNG device' seed mode, using '/dev/urandom', seed=-1066432186 seed_offset=0 real_seed=-1066432186
basic.random.init_random_generator: RandomGenerator:init: Normal mode, seed=-1066432186 RG_type=mt19937
core.init: Resolved executable path: /media/sajad/18a2b4cf-06c3-437d-bf57-7d68790182b9/sajad/Desktop/Tools/Lab/Modeling/Rosetta/rosetta_bin_linux_3.13_bundle/rosetta_bin_linux_2021.16.61629_bundle/main/source/build/src/release/linux/5.13/64/x86/gcc/9/default/mp_span_from_pdb.default.linuxgccrelease
core.init: Looking for database based on location of executable: /media/sajad/18a2b4cf-06c3-437d-bf57-7d68790182b9/sajad/Desktop/Tools/Lab/Modeling/Rosetta/rosetta_bin_linux_3.13_bundle/rosetta_bin_linux_2021.16.61629_bundle/main/database/
apps.public.membrane.mp_span_from_pdb: spanfile_from_pdb
core.chemical.GlobalResidueTypeSet: Finished initializing fa_standard residue type set. Created 984 residue types
core.chemical.GlobalResidueTypeSet: Total time to initialize 0.501106 seconds.
core.import_pose.import_pose: File 'mp_ddg/inputs/MC4R_Inactive_model_withNA_A.pdb' automatically determined to be of type PDB
core.io.pose_from_sfr.PoseFromSFRBuilder: [ WARNING ] PDB reader is ignoring atom OXT in residue 278 A. Pass flag -ignore_zero_occupancy false to change this behavior
core.conformation.Conformation: [ WARNING ] missing heavyatom: OXT on residue ILE:CtermProteinFull 278
core.conformation.Conformation: Found disulfide between residues 1 240
core.conformation.Conformation: current variant for 1 CYS
core.conformation.Conformation: current variant for 240 CYS
core.conformation.Conformation: current variant for 1 CYD
core.conformation.Conformation: current variant for 240 CYD
core.conformation.Conformation: Found disulfide between residues 232 238
core.conformation.Conformation: current variant for 232 CYS
core.conformation.Conformation: current variant for 238 CYS
core.conformation.Conformation: current variant for 232 CYD
core.conformation.Conformation: current variant for 238 CYD
apps.public.membrane.mp_span_from_pdb: Taking default thickness: 30
protocols.DsspMover: LLLLLLLLHHHHHHHHHHHHHHHHHHHHHHHHLLHHHLHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHLLLLLLHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHLLHHHHHHHHHHHHHHHHHHHHHHHHHHHLHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHLLLLLLLLLLLLLLLLLHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHLLLLHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHLHHHHHHHHHHL
core.conformation.membrane.SpanningTopology: create topology from structure
core.conformation.membrane.SpanningTopology: [ WARNING ] No spans calculated from structure for chain 1

ERROR: Can't print SpanningTopology. It's empty!
ERROR:: Exit from: src/core/conformation/membrane/SpanningTopology.cc line: 167

[ ERROR ]: Caught exception:

File: src/core/conformation/membrane/SpanningTopology.cc:167
[ ERROR ] UtilityExitException
ERROR: Can't print SpanningTopology. It's empty!

AN INTERNAL ERROR HAS OCCURED. PLEASE SEE THE CONTENTS OF ROSETTA_CRASH.log FOR DETAILS.

Attachment	Size
MC4R_Inactive_model_withNA_A.pdb	168.17 KB

Category:

Post Situation:

Forums:

Unable to locate the "compute_ddg.py" file and running the RosettaMP protocol

Sajjad — Mon, 28 Nov 2022 22:09:10 +0000

Dear users,

I am trying to run the RosettaMP application to predict the effect of several mutations on the stability of a membrane protein. I am following this documentation to do that :"https://www.rosettacommons.org/docs/latest/application_documentation/membrane_proteins/RosettaMP-App-MPddG"

However, I can not locate the "compute_ddg.py" in the expected derectory (source/src/python/bindings/app/membrane/compute_ddG.py).

There is also no binary version of this tool in /main/source/bin directory.

Could you please give me a clue on how to resolve this issue?

Thank you

Category:

Forums:

Post Situation:

rosetta scripts to superimpose multiple segments

rlwoltz — Fri, 29 Apr 2022 22:15:09 +0000

Dear community,

BACKGROUND:

I am homology modeling an ion channel and I'm struggling with keeping the selectivity filter of the channel from moving. I've done a lot of trouble shooting and near the end but have one last hiccup. The current step I'm on is a final relax using franklin2019 from each of the 1000 decoys outputted by rosetta homology modeling. Since the oxygens in the selectivity filter are so repulsive the filter is expanding by a significan amount. However, if I include K+ in the filter it preserves much better. So I need to find a way to automate placing K+ in very specific locations inside the selectivity filter and they need to be centered inside the 8 oxygens that make up a cube or cage for the ion. This needs to be automated because I have 1000 models to relax and it would take forever to do this manually. My solution is to create a single decoy with the appropriate ion placement then using the superimpose mover align the remaining decoys using the helices that make up the pore domain or selectivity filter, then copy and paste the coordinates from the potassium atoms. I've used this method to extract Ca2+ ions into other channels but the region I was aligning was a single continous segment.

The Problem

The residue numbering of the pore domain helices is 193-277,718-802,1243-1327,,1768-1852. If I align just one segment (193-277) or include all residues inbetween (193-1852) we have problems of favoring one side and the structure being slightly tilted or the flexible that were remodeled loops inbetween the segments pulling the structure aways from a good alignment.

Is there a way to align multiple sections of a protein that are not connected simultaniously? i.e:

9: <MOVERS>
10: <Superimpose name="align" ref_start="193,718,1243,1768" ref_end="277,802,1327,1852"
11: target_start="193,718,1243,1768" target_end="277,802,1327,1852" ref_pose="hsk2-cam-homology-from-map-refined-6cno-alignment-template-1-all-chain-A.pdb" CA_only="1"/>
12: </MOVERS>

or to make multiple movers and have teh superimposition averaged in the protocol section?

Is there another program I could use (VMD/Chimera) which I could automate with a script to do this as I know I can command line align multiple selections through those programs but there is a level of user interface and waiting for program to load I don't know how to script around.

Any directions would be super helpful thank you!

Ryan

Category:

Forums:

Post Situation:

Relax in membrane pull my protein out of the membrane

Martin Floor — Wed, 26 Jan 2022 15:46:05 +0000

Good day,

I am having issues with the mp_relax protocol. I am testing the protocol directly from a PDB crystal structure, and the relax algorithm takes this structure out from the membrane region (as defined by the MEM residue). I am first optimizing the membrane location with the mp_transform (with the -mp:transform:optimize_embedding true option), which positions the protein in the membrane as expected. Then, the relax protocol pulls out the protein from this "correct" membrane position, leaving the protein completely exposed to solvent.

My XML file is:

<ROSETTASCRIPTS>
  <SCOREFXNS>
    <ScoreFunction name="mpframework_smooth_fa_2012" weights="mpframework_smooth_fa_2012"/>
  </SCOREFXNS>
  <RESIDUE_SELECTORS/>
  <JUMP_SELECTORS/>
  <TASKOPERATIONS/>
  <MOVE_MAP_FACTORIES/>
  <SIMPLE_METRICS/>
  <FILTERS/>
  <MOVERS>
    <AddMembraneMover name="addMembraneMover"/>
    <MembranePositionFromTopologyMover name="init"/>
    <FastRelax name="fastRelax" repeats="5" scorefxn="mpframework_smooth_fa_2012"/>
  </MOVERS>
  <PROTOCOLS>
    <Add mover="addMembraneMover"/>
    <Add mover="init"/>
    <Add mover="fastRelax"/>
  </PROTOCOLS>
  <OUTPUT scorefxn="mpframework_smooth_fa_2012"/>
</ROSETTASCRIPTS>

My fastrelax and membrane relevant options are:

-ex1
-ex2
-use_input_sc
-flip_HNQ
-no_optH false
-mp::setup::spans_from_structure true

I am puzzled since the mp_transform finds a good-enough position, however, "relax" does not like it. Does anyone have a hint on what could be going on here?

Attachment	Size
Trajectory showing the mp_relax behaviour	373.93 KB

Category:

Post Situation:

Forums:

calcium metal nomenclature: Rosetta_cm confusing HETATM CA (calcium) with ATOM CA (alpha-carbon)

rlwoltz — Wed, 04 Aug 2021 06:26:29 +0000

Dear community,

I had this posted under a different sub-forum but I realized it didn't fit that forum so deleted it and reposting here.

Hopefully this is a simple answer but I couldn't find any answer and I don't remember this happening on a previous rosetta version, just updated from 2019 to 2021.16.61629_bundle. The goal is to add constraints to the amino acids that bind calcium in a homology model since they are in a small loop and can reorient without calcium present. I'm using the metal_setup mover to apply constraints but I'm open to other suggestions.

I'm doing a homology model of a 4-unit symmetrical membrane channel coupled with Calmodulin. The channel and Calmodulin are seperate chains which is denoted in the fasta with a "/" in the sequence to show the split. All this works fine for un-calcified calmodulin. However, now I'm making a calcified calmodulin structure and I have calcium ions in the single unit INPUT.pdb (output of make_symm.sh file. The thread is made with no issue. The problem is the next step of using rosetta_CM and the xml. I'm using the metals setup mover line (1) to denote the amino acids and residues. However, the calcium ions are being confused by roestta for Alpha-Carbons. I have three Calciums per subunit and they are all being overlayed onto the last alpha carbon in the protein sequence for each single unit (2).

Is there a special residue or atom name the rosetta is expecting for Calciums? currently the only naming scheme that works is "CA". (3)

Thanks for any help and let me know if you need any more information,

Ryan

1. calcium is residues 526-528. interacting residues is 397-408 --> 526, 433-433 --> 527, 470-480 --> 528

2. notice that the Calciums have the same coordinates as the first alpha-carbon of chain B the start of calmodulin. this is true for all 4 symmetrical units

ATOM 6185 OD1 ASN A 379 -2.989 8.407 56.731 1.00 0.00 O
ATOM 6186 ND2 ASN A 379 -3.678 8.654 54.615 1.00 0.00 N
ATOM 6187 H ASN A 379 -5.605 5.733 55.922 1.00 0.00 H
ATOM 6188 HA ASN A 379 -4.757 7.483 58.184 1.00 0.00 H
ATOM 6189 1HB ASN A 379 -5.936 8.097 55.462 1.00 0.00 H
ATOM 6190 2HB ASN A 379 -5.666 9.317 56.708 1.00 0.00 H
ATOM 6191 1HD2 ASN A 379 -2.719 8.727 54.264 1.00 0.00 H
ATOM 6192 2HD2 ASN A 379 -4.442 8.704 53.978 1.00 0.00 H
TER
ATOM 6194 N ASP B 380 -1.237 -47.247 33.849 1.00 0.00 N
ATOM 6195 CA ASP B 380 -2.006 -46.260 34.601 1.00 0.00 C
ATOM 6196 C ASP B 380 -2.779 -45.354 33.619 1.00 0.00 C
ATOM 6197 O ASP B 380 -3.335 -45.841 32.634 1.00 0.00 O
ATOM 6198 CB ASP B 380 -1.035 -45.467 35.514 1.00 0.00 C
ATOM 6199 CG ASP B 380 -1.668 -44.574 36.608 1.00 0.00 C
ATOM 6200 OD1 ASP B 380 -2.103 -44.975 37.659 1.00 0.00 O
...

HETATM 8425 CA CA C 526 -2.006 -46.260 34.601 1.00 0.00 CA
HETATM 8426 CA CA C 527 -2.006 -46.260 34.601 1.00 0.00 CA
HETATM 8427 CA CA C 528 -2.006 -46.260 34.601 1.00 0.00 CA

3. last few lines of input file. notice no TER line although I've tried it with. notice different XYZ coordinates for all atoms. I've also tried adjusting the spacing of the "CA" in the residue column. Calmodulin for the input as chain A but converted to chain B due to the fasta file input described above.

ATOM 5840 CA ASP A 380 -5.804 -51.544 37.679 1.00 0.00 C
...

ATOM 8068 3HB ALA A 525 -34.729 -39.328 36.187 1.00 0.00 H
HETATM 8069 CA CA A 526 2.182 -32.441 49.215 1.00239.49
HETATM 8070 CA CA A 527 -8.813 -37.633 50.564 1.00182.89
HETATM 8071 CA CA A 528 -33.639 -23.142 47.008 1.00223.54

Category:

Forums:

Post Situation:

FastRelax Mover with symmetry, ligand and membrane

Michele.Bonus — Wed, 26 May 2021 11:43:44 +0000

Dear all,

I am currently trying to run the FastRelax Mover in RosettaScripts to relax a C4-symmetric protein with 4 ligands (1 per monomer). My input structure is a single protomer and a ligand (the pdb file unfortunately exceeds the maximum file size). My xml and option files look as follows. Note that I don't actually use the MembranePositionFromTopologyMover because it fills up 16 GB of memory pretty quickly. This leads to my first question. Can I just omit this mover if the protein is reasonably aligned and I specify a spanfile in the (Symmetric)AddMembraneMover?

XML File:

<ROSETTASCRIPTS>
    <TASKOPERATIONS>
        <RestrictToRepacking name="restrict"/>
        <IncludeCurrent name="keep_curr"/>
    </TASKOPERATIONS>
    <SCOREFXNS>
        <ScoreFunction name="memb_hires" weights="%%sfxn_weights%%"/>
    </SCOREFXNS>
    <FILTERS>
    </FILTERS>
    <MOVERS>
        <SetupForSymmetry name="setup_symm" definition="mHCN2_from_6UQG_07_S_0001.symm" set_global_symmetry_at_parsetime="0"/>
        <SymmetricAddMembraneMover name="add_memb" spanfile="mHCN2_from_6UQG_A.span"/>
        <MembranePositionFromTopologyMover name="init_pos"/>
        <FastRelax name="fast_relax" scorefxn="memb_hires" repeats="8"/>
    </MOVERS>
    <APPLY_TO_POSE>
    </APPLY_TO_POSE>
    <PROTOCOLS>
        <Add mover="setup_symm"/>
        <Add mover="add_memb"/>
        <!--Add mover="init_pos"/-->
        <Add mover="fast_relax"/>
    </PROTOCOLS>
    <OUTPUT scorefxn="memb_hires"/>
</ROSETTASCRIPTS>

Options file:

-in:file:s mHCN2_from_6UQG_07_S_0001_INPUT.pdb
-in:file:extra_res_fa ../04_Threading/04_Ligand_Parameters/CMP.fa.params
-parser:script_vars sfxn_weights=franklin2019
-mp:setup:spanfiles mHCN2_from_6UQG_A.span
-parser:protocol mHCN2_from_6UQG_relax.xml
-out:file:scorefile model_scores.sc
-mp:lipids:composition DLPC
-out:pdb
-relax:jump_move true
-relax:fast
-packing:pack_missing_sidechains false
-score:weights franklin2019

In any case, RosettaScript exits with a rather uninformative error unless I delete the ligand from the PDB file.

AN INTERNAL ERROR HAS OCCURED. PLEASE SEE THE CONTENTS OF ROSETTA_CRASH.log FOR DETAILS.


terminate called after throwing an instance of 'std::out_of_range'
  what():  map::at
Got some signal... It is:6
Signal 6 (SIGABRT) means that the process was aborted.  This usually means an internal Rosetta error caused by (often) bad inputs, (sometimes) developer error, or (rarely) hardware problems.

Next, I figured, I should add the Ligand to the pose using the AddMPLigandMover, so I changed my XML file accordingly:

[...]
    <MOVERS>
        <SetupForSymmetry name="setup_symm" definition="mHCN2_from_6UQG_07_S_0001.symm"/>
        <SymmetricAddMembraneMover name="add_memb" spanfile="mHCN2_from_6UQG_A.span"/>
        <AddMPLigandMover name="add_lig" ligand_seqpos="1" closest_rsd="450"/>
        <MembranePositionFromTopologyMover name="init_pos"/>
        <FastRelax name="fast_relax" scorefxn="memb_hires" repeats="8"/>
    </MOVERS>
    <APPLY_TO_POSE>
    </APPLY_TO_POSE>
    <PROTOCOLS>
        <Add mover="setup_symm"/>
        <Add mover="add_memb"/>
        <Add mover="add_lig"/>
        <!--Add mover="init_pos"/-->
        <Add mover="fast_relax"/>
    </PROTOCOLS>
    <OUTPUT scorefxn="memb_hires"/>
</ROSETTASCRIPTS>

Unfortunately, RosettaScripts didn't like the idea of adding a ligand to an already symmetric membrane pose. It seems the AddMPLigandMover can only add ligands to non-symmetric membrane poses (I checked this on the monomer and it worked).

In thinking of a workaround, I decided to add a membrane to the monomer pose first, add the Ligand with the AddMPLigandMover, then SetupForSymmetry and then run FastRelax. This worked, the protein looks ok (despite a lot of "inaccurate G! step" warnings) and the ligand is there in all four protomers. Here is the final working XML file:

<ROSETTASCRIPTS>
    <TASKOPERATIONS>
        <RestrictToRepacking name="restrict"/>
        <IncludeCurrent name="keep_curr"/>
    </TASKOPERATIONS>
    <SCOREFXNS>
        <ScoreFunction name="memb_hires" weights="%%sfxn_weights%%"/>
    </SCOREFXNS>
    <FILTERS>
    </FILTERS>
    <MOVERS>
        <AddMembraneMover name="add_memb" spanfile="mHCN2_from_6UQG_A.span"/>
        <AddMPLigandMover name="add_lig" ligand_seqpos="580" closest_rsd="450"/>
        <SetupForSymmetry name="setup_symm" definition="mHCN2_from_6UQG_07_S_0001.symm"/>
        <FastRelax name="fast_relax" scorefxn="memb_hires" repeats="1"/>
    </MOVERS>
    <APPLY_TO_POSE>
    </APPLY_TO_POSE>
    <PROTOCOLS>
        <Add mover="add_memb"/>
        <Add mover="add_lig"/>
        <Add mover="setup_symm"/>
        <Add mover="fast_relax"/>
    </PROTOCOLS>
    <OUTPUT scorefxn="memb_hires"/>
</ROSETTASCRIPTS>

Since i symmetrize after adding the membrane, I now have 4 MEM residues that seem slightly different (see attachment). Although the protein looks OK, I have doubts about the results because of this. Is there any "non-workaround" solution that I should clearly be using here? I couldn't find any AddMPLigandMover equivalent that works on symmetric poses.

I would appreciate any help!

Best regards

Michele

Attachment	Size
1.jpg	315.34 KB
2.jpg	208.93 KB

Category:

Post Situation:

Forums:

MPI optimization on TACC stampede2 HPC

rlwoltz — Wed, 17 Mar 2021 01:07:26 +0000

Dear community,

I am new to MPI so please correct any misuse of terms. My question is about optimizing for MPI on an HPC or a second (preferred) solution to my issue is to optimize my RAM usage so I don't have to run MPI, serial would actually improve overall models produced anyways.

First I have a very simple question, in the output of rosetta how do I know how many cores are being used in the MPI? I see (0), (1), and (2) at the beginning of each line so I'm assuming this mean 1 2 and 3 cores being used for that output line, is this correct? Basically if I'm using 2 cores with MPI in the output how do I know they actually got used?

I'm trying to optimize my rosetta MPI run on an HPC and my code currently is running 3X slower than expected and from reading documentation I think it just might be an option incompatibility so hoping someone with experience can advise. I have a lot of experience running with HPCs charging by the CPU but I'm using stampede2 which charges by the node so things get complicated here. Normally I'd prefer to run completely in serial but the maximum RAM the node has is 96GB and my system uses 3GB per run limiting me to about 35 runs per node (emperically tested this) but I have 68 available cores thus the want to include MPI to take advantage of the 36 idle cores (assuming I play it safe and only run 32 runs/node). Again with the HPC I'm using you can only request an entire node not individual CPUS.

So this is my plan of attack:

Request 1 node which has 96GB of RAM and 68 Xeon PHI CPUs. run 32 independent jobs to prevent exceeding RAM but each job will have 2 CPUs working in MPI.

programs I'm using to submit (screenshots attached):

I have 4 scripts that pass things along. First I have my sbatch submission script which uses the launcher module. The main objective in this is to request 32 tasks to be run at a time with 2 CPUs per task. Then I have a jobfile that essentially lists the total number of commands (tasks) I'd like to run, for example 320 tasks would run 10 rosetta runs on 32 (X 2 for MPI) CPUs in parallel and independently. This job file runs the rosetta command file which must use mpiexec (not mpirun) command explained below. Finally this also calls the rosetta.xml but this doesn't really play a role in the sbatch but it is there for completeness.

Double checking run:

When checking top on the node the job is submitted to 65 CPUs are in use (64 for rosetta jobs and 1 for node managment?). However, I do notice that there are 2474 total, 61 running and 2413 sleeping so not sure what this refers to maybe threads (NOTE: these numbers are for my latest 30 CPU test not 32 as I'm describing)? Unfortunately, top is the only montoring program (I use htop) so if there is a command that can be used to officially check the CPUs being used I'd appreciate it. also in the rosetta output it has a (0) and a (1) for each line. I'm assuming this refers to 1 or 2 cpus being used but please clarify if I'm wrong. When I tested with 4 CPUs the max number I saw was (2) so maybe a solution is to run 22 tasks and MPI CPU=3?

output files (don't think it is helpful for troubleshooting but might be helpful in advising where to ask me to start looking):

Jobfile also has the rosetta terminal output saved with a file embedded with a unique stampede2 job id, launcher job # (1-320), and a launcher task id which I believe is the CPU id number running the script. Each Rosetta output silent files also has these unique embedded codes and saves all silent files to unique results directory.

The problem:

So everything runs and I believe it is doing MPI in the way that I'd like it to but don't know how to check other than the (x) line mentioned above. But it is very ineffecient and takes about 13 hours versus 4-6 hours per model on my laptop. I'm doing a homology model with symmetry, membrane, and loops of a protein that is 2100 AA in total and I've spent a lot of time optimizing this part. My preference obviously is to run completely in serial but because of RAM limitations this isn't possible. I'd assume the MPI would be faster, not 3X slower, than serial so I'm just wondering if I can optimize the options here. I read this about "bind to none" option in openmpi but again I've never dealt with MPI and so the terminaology is difficult for me to understand. Again in the top there are a lot of sleeping tasks so this is why I think it could be optimized. I've attached screenshots of the slurm submission script, the jobfile script, and the mpiexec rosetta run script.

Possible solutions:

I've played around with options but because of the queue and the 13 hours to make a model it is taking days-weeks to get any meaningful results. I did test with a shorter 5 min run but it didn't scale. I'm continueing to try new things but thought I could save time by asking for advice. I have a feeling it has something to do with either the launcer slurm script or the way I'm calling for MPI. The launcher slurm script calls for -n = 32 # number of tasks ---cpus-per-task=2. In the rosetta MPI running file I have this line:

mpiexec.hydra -genv I_MPI_FALLBACK=1 -n 2 rosetta_scripts.cxx11mpi.linuxiccrelease \

-genv I_MPI_FALLBACK=1 #must be included otherwise the job fails. Something with Fabric and fallback disabled so don't understand this bit but included because it works.

-n 2 #number of cpus per task? but the document I read suggested this is how many time the command is run so I'm confused by this, maybe it should be 32?

There is also this passage:

Please note that mpirun automatically binds processes as of the start of the v1.8 series. Three binding patterns are used in the absence of any further directives:

Bind to core:: when the number of processes is <= 2
Bind to socket:: when the number of processes is > 2
Bind to none:: when oversubscribed; It sounds like I should bind to core since I'm running with 2 but really I'm running with 32X2 so should I bind to none?; MPIrun document:; https://www.open-mpi.org/doc/v3.0/man1/mpirun.1.php; Sorry if this isn't a problem so much as helping me improve my code and unfortunately no one in my network has experience running rosetta in MPI.

Thank you for your help and please let me know if more information or clarification is needed,

Ryan

Attachment	Size
launcher slurm	287.08 KB
jobfile launcher	491.5 KB
rosetta mpi script	310.21 KB

Category:

Post Situation:

Forums:

core.optimization.LineMinimizer: [ ERROR ] Inaccurate G!

aralz — Thu, 11 Feb 2021 08:46:23 +0000

I am doing an efficiency analylis on the different minimization options in Rosetta using FastRelax, MPRelax and lbfgs_armijo_nonmonotone minimizer on a membrane protein using the franklin2019 score function and I very very often get this error:

core.optimization.LineMinimizer: [ ERROR ] Inaccurate G!

It is clear to me that they do not perform as well when the error occurs but how worried should I be about it? I can get FastRelax and armijo to work without giving the error but MPRelax always has a bunch of those. I have read that it only worsen the accuracy but it is still useful, is that correct?

Thank you.

Category:

Forums:

Post Situation: