# The confussion about an example in tutorial in Rosetta Virtual Workshop 2020

4 posts / 0 new
The confussion about an example in tutorial in Rosetta Virtual Workshop 2020
#1

Dear all,

I am a new user of rosetta software.In order to get familiar quickly with this software I plan to repeat the tutorial in Rosetta Virtual Workshop 2020 ,following you guys to have a better understanding of every application in this  versatile software.But at the very beginning it is very hard for a layman to know every point.I am stuck in the first example called protein_folding for nearly one week.Along this way I overcame many difficulties by myself .But when it come to some knowledge as to computer,I have no idea.Now I just want to solve a problem which I can not find any solution to  it .This question is that when I want to perform the command "~/rosetta_workshop/rosetta/main/source/bin/AbinitioRelax.default.linuxgccrelease @2LZM_broker.options"(the path I actually use is correct),initially i am not adding any parameter as a result it take nearly whole day and can even not finish one step ,stucking in "protocols.jobdist.JobDistributors: (0) Master Node -- Waiting for job request; tag_ = 1".After that I try another node with 32 processors adding the mpi paremeters.The whole command line is that "mpirun -np 32 AbinitioRelax.mpi.linuxgccrelease @2LZM_broker.options".Consequently it generates a lot of "core.*"files and shows that:

basic.io.database: (28) Database file opened: scoring/score_functions/P_AA_pp/P_AA
basic.io.database: (28) Database file opened: scoring/score_functions/P_AA_pp/P_AA_n
core.scoring.P_AA: (28) shapovalov_lib::shap_p_aa_pp_smooth_level of 1( aka low_smooth ) got activated.
basic.io.database: (28) Database file opened: scoring/score_functions/P_AA_pp/shapovalov/10deg/kappa131/a20.prop
protocols.jobdist.JobDistributors: (28) Node: 28 next_job()
protocols.jobdist.JobDistributors: (28) Slave Node 28 -- requesting job from master node; tag_ 1
protocols.jobdist.JobDistributors: (0) Master Node -- Spinning down node 28 with 2 remaining nodes.

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

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

protocols.abinitio: (23)
------------------------------------------------------------
Scores                       Weight   Raw Score Wghtd.Score
------------------------------------------------------------
vdw                          1.000       0.436       0.436
pair                         1.000      -3.147      -3.147
env                          1.000     -16.491     -16.491
hs_pair                      1.000       0.000       0.000
ss_pair                      0.300       0.000       0.000
sheet                        1.000       0.000       0.000
---------------------------------------------------
Total weighted score:                      -19.202
protocols.moves.TrialCounter: (23) ClassicFragmentM trials=  20000;  accepts= 0.3760;  energy_drop/trial=  -0.00047
protocols.abinitio: (23)
===================================================================
Stage 3
Folding with score2 and score5 for 20000

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

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

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

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

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

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

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

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

--------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------

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

--------------------------------------------------------------------------
mpirun noticed that process rank 17 with PID 0 on node d007 exited on signal 11 (Segmentation fault).

The ROSETTA_CRASH.log I attched at the end of this mail.

The magic thing is that when I change the number of the parameter"np" from 32 to 2,it works and finished in just about 5mins! Only this number can work,if the number is 1,it come back to the problem I have told above and until now I have no idear what is the result if it finish running(it still running now).And I also have no idea whether the number 1 and 2 can generate the same results. If the results are diffenrent ,which is right?

Everything is the same with the tutorial but the command line.

I would really appreciate it if you help me to understand this particular case.

AttachmentSize
83.94 KB
Category:
Post Situation:
Tue, 2022-11-29 04:58
ldx022

I don't have experinence with folding, so I only have two comments:

• Ab initio folding is not that useful any more now that AlphaFold2 is generally available.
• You are using an old version of Rosetta (2021.04) perhaps try with a newer version?

Tue, 2022-11-29 06:00
ajasja

Theoretically, you should be able to run the MPI-compiled version of Rosetta without the mpi launcher as a single threaded job. But that doesn't always work. As you've seen, sometimes it just stalls and doesn't do anything, waiting for other processes to join. Your best bet is to use the MPI compile for MPI runs, and a non-MPI compile for non-MPI runs.

From the ROSETTA_CRASH.log, my best guess at what is going wrong is that you're running MPI with 32 processors, but only asking for a single output structure (-nstruct 1 in the options file.) Most Rosetta protocols (including AbInitio) are "trivially parallelizable", which effectively means that each output structure is assigned to a single CPU, and throwing more CPUs at it won't help. More CPUs *are* helpful in the general case where you have hundreds to thousands of output structures -- each CPU gets its own output and chugs away at it independently of the other processors. But if you only have one output, then the other processors are standing around with nothing to do. This shouldn't be a problem (the other processors should just idle while you finish the rest), but there may be a bug when assigning the initial set of jobs.

That said, I agree with ajasja that AbInitio is probably not all that useful these days, unless you're doing something highly specialized. (The 2020 workshop is the last workshop at which there will be an AbInitio component.) If you're just looking for a standard sequence->structure prediction, AlphaFold or RoseTTAFold (or whatever new hotness comes out of this year's CASP) is probably a better bet.

Tue, 2022-11-29 07:26
rmoretti

Thanks a lot!!!!!!!!Your suggestions really really hit me.I really  made such a stupid mistake .So it is much more important to learn to utilize the advanced methods or tools according to the requirements rather than insisting on the outdated tools and method .But what advantages do the rosetta have while others don’t?

Just as I wrote before ,rosetta is the first software I have ever used to deal with the protein problem .The reason why I want to learn this software is that my advisor wants me to achieve the transformation of enzymes to improve its activity .Such as through the mutation site of enzymes to improve the activity of some reactiont or for other purpose.Based on this idea,I wonder if rosetta can help me out in some way but it seems to me it has so much application so that I have a long way to explore .It  confused me for nearly one month.