The Zhang Lab Message Boards

Hello!
I am new to using I-Tasser, I would like to send a protein that has a crystallographic structure in the pdb database to model a part that interests me. I want to use "Specify model without alignment", but putting two strings. How could I guide modeling in two chains?

I manipulated my amino acid sequence into a text box and got an optimal model(http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S555884/),save as PDB. The evaluation data of this model :C-score=-0.72,TM-score=0.62±0.14,RMSD=6.3±3.9A;Then, I put this model into SAVES v5.0(https://servicesn.mbi.ucla.edu/SAVES/) to do an assessment, the output shows a problem:have no END tag, see the attached file for details. I want to know whether the "no END tag" means that the model is an incomplete protein? Have you ever had a problem like this before?
Looking forward to your reply!

Hello,

I am running ITASSER Suite on a computing network and it never gets past step 4a "run I-TASSER simulation". Before it even runs the first simulation, it times out. I have tried running one node with 24 processors for twenty hours and on two nodes with 48 processors for four hours. The output file reads:

"1. make seq.txt and rmsinp
Your protein contains 283 residues:
> Protein
MALKDDAVLIAARGYVYTAAVGTAAPTPSQLKLIDLEHPEAWDRTGWDLVGHTSEDDLPE
FGFDGGDSEVRGSWQKKKLREVETEEIADYVVINLTQFDESALELYFGPNQSATPGIFGV
KSGSVVNERALLIVIVDNDVRLGFHARKASLKREDAISLATDEFGALPVRATFLDYQSYN
LYEWIEEDWFNAADAPVVYLLDLGGATGGDYTLLVGGKSTGDIAYNANASAIKTAIGAVD
DGVAESAWTVTADGSDFEISGPLAVALGVDSTTGGSGVTVDVA
2.1 run Psi-blast
2.2 Secondary structure prediction was done before.
2.3 Predict solvent accessibility...
2.4 run pairmod
pair exist
3.1 do threading
/home/ek/ProteinSeq/Protein/init.PPAS exists
/home/ek/ProteinSeq/Protein/init.dPPAS exists
/home/ek/ProteinSeq/Protein/init.dPPAS2 exists
/home/ek/ProteinSeq/Protein/init.Env-PPAS exists
/home/ek/ProteinSeq/Protein/init.MUSTER exists
/home/ek/ProteinSeq/Protein/init.wPPAS exists
/home/ek/ProteinSeq/Protein/init.wdPPAS exists
/home/ek/ProteinSeq/Protein/init.wMUSTER exists
3.2 make restraints
4.1 run simulation
run 14 parallel simulations
run the first simulation job Proteinsim_1A"

I'm not sure if I'm doing something wrong. My command line is: "runI-TASSER.pl -libdir /depot/mylab/itasserlib -seqname Protein -datadir /home/ek/ProteinSeq/Protein".

Should I just keep trying for longer amounts of time? Or is there likely something wrong with my code?

Dear scholar,
hello！I have some problems when running ResPRE.
I run the following command:
python respre.py test/T0759-D1.aln test/T0759-D1.out 、
Then the console displays:
--------------------------------------------------------------------------------------
cuda is ready? : True
/media/ubuntu/619b7769-bf19-45af-9e3e-90232241302c/home/ubuntu/terence/ResPRE/precision.py:40: UserWarning: genfromtxt: Empty input file: "test/T0759-D1.out.weight"
weights=np.genfromtxt(weightfile).flatten()
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [100%]
--------------------------------------------------------------------------------------
I have 2 questions：
1.What is the effect of this file （T0759-D1.out.weight）?
2.It reminded that file T0759-D1.out.weight was Empty input file. Does this affect the prediction results?
Hope you can give me the answer, thank you very much!

While i am running the program, it seems PDB file cannot be opened. Any suggestions to resolve this?
Also i don't have any reference protein, so will there be problem while using the same protein as reference?

$ ./mcrefinement Desktop/ModRefiner Desktop/ModRefiner c_mutant.pdb c_mutant.pdb 0
PDB file can not be opened Desktop/ModRefiner/c_mutant.pdb

Hello

I am using I-Tasser to predict the structure of a protein. At step 5.1 I get the following error:
"
5.1 do clustering
No. of trajectory files: 62
Calculating RMSD scores
8.00000000 3.50000000 12.0000000
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO
cp: cannot stat './closc*.pdb': No such file or directory
cp: cannot stat './combo*.pdb': No such file or directory
5.2 build full-atomic model
"
What should I do?

Any help is appreciated in advance
Mahdi

I keep getting the following error.
" We have received 1 jobs submission from this IP address, which are currently pending/running."

To the best of knowledge I do not have outstanding job submission. Can you remove any pending jobs or give me instructions about how to do so, so that I can submit a new job.

Thanks,

Andrea
adooley@med.umich.edu

I was trying to use SSIPe standalone program. However something is going wrong. I'm preparing the "complex.pdb" and "mutList.txt" files and running the following code;
../SSIPe.2019-11-09/bin/get_final_score3.py ./

and then following error is printed in stdout;
step 4: get_evoef_score
mv: target '.pdb' is not a directory
EvoEF failed generating wild-type or mutant model for mutant VA471G

What is wrong here? Is program failing to create mutant pdb files or should I create them in someway beforehand? If second, how can I create those mutant complex ? There are hundreds of mutations in the list.

I submitted a sequence for structure modeling on June 12, 2020 (https://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S549093/). Its status has remained as “This job is running and should be completed in approximately 60hrs.“ at least for the past two weeks. Besides, all jobs submitted around that time was completed. Could you please check what the problem is? The sequence is 506 AA long and so it is not huge.

Thanks very much in advance!

To all it may concern,

My name is Matthew and I am trying to dock a sialoglycoprotein(CD24) with another protein for which I already have a decent homology model. I generated a model of the protein core of CD24 in I-TASSER, and was planning to add carbohydrates to the model later. However, the C score for the model was -2.4, and I am not sure if it is suitable for docking as I don't have much prior experience in protein modeling. I could really use an expert's opinion on whether or not my current model of CD24 will help indicate how it binds to other proteins in vitro. If the model is not sufficient, then I would really appreciate advice on how to create a more accurate model.

Thanks,
Matthew

Hi.

I submited a work 3 weeks ago and i want to cancel it, but i didnt get the email with the ID.

Ty.

Hi,

I have a pending i-tasser job "BIIICH1withNT" in queue that has a submission error and needs to be removed. This job is private, but I did not receive an email with a pass code to access/remove the job. Currently, until this job is removed, I cannot submit a new job with the submission correction. What do I do?

Adam

I don't think I do have a job running. I think my last job was S546905. It finished but I tried to cancel it anyways, didn't work....

If I do have a job running, can you let me know the ID so I can cancel it? And if I don't please allow me to run a job.

Thank you!

-Alan

Dear All,

I would be really grateful if someone can help in predicting/analyzing the domain-domain interaction within the query protein using the i-TASSER result.

On the 18th of June I submitted an I-TASSER job (ID: S550449) which was approximated to be ready in about 35 hours. However, the job hasn't been completed yet and is still 'in processing'. I used I-TASSER in the past and it never took this long, is it because of the server overload? At the same time jobs submitted at the same time as mine and after are completed normally. Thank you.

I set my protein up weeks ago and the estimated time has not shifted from 36 hours. Why does this put it in an infinite loop rather than just eventually giving up and letting me resubmit? Is this an issues for anybody else? I think the issue was that I set up for a template without alignment using a pdb ID and specifying the chain. However, I-TASSER won't allow me to cut my losses.

Dear developers,

I am running the standalone I-TASSER v5.1 on a shared computer cluster. I typically request 10 CPUs and a single node, and run the software in the Linux system with the these extra parameters: -homoflag real -light true -runstyle gnuparallel.

I was reported by the administrators that there were 7000 file opens and closes in the span of 10 seconds for the job I submitted. And it seems this consistently happens in all the jobs I submitted for I-TASSER. Is it a typical part of the pipeline?

Dear Sir/Mam,

Since I am new to i-TASSER, i would like to ask how much time will it take to complete the job id S549595?

I run I-TASSER 5.1 with -runstyle gnuparallel and there are 14 gnuparallel simulations.
How to change simulations runs number ?
thank you

Dear Zhang Lab Service System,
My professor has already sent an email, but I thought I would send one to indicate the importance of the message. I noticed that job id S546873 is not finished yet and I really need it as soon as possible for research purposes and I was wondering if you can possibly notify us when it will be done, thank you and hope to hear from you soon.

Steffany

Hello,

My student submitted to I-tasser a protein structure prediction job (id S546873) on June 3rd. On the queue list, we read the message "This job is running and should be completed in approximately 35hrs". Could you please provide an estimate of when the job will finish?

In the past, my student has submitted jobs that finish within one week.

Dear I-TASSER team,

We ran two modeling jobs on I-TASSER ~three weeks ago (21.05.2020), but it seems that both are still in processing, while most of jobs submitted around that time, and also more recent ones, have already been processed.
Usually when we ran I-TASSER jobs, they were finished after 2-3 days, but, as my student reports, the unusually long waiting time happened to him as well during last 2 months.

Below are job IDs of current runs suffering from this problem:
S543353
S543200
Both of these are small proteins (~120aa) with provided template pdb (no alignment) and secondary structure assignment for 5 amino acids.
An another job that I ran recently, regarding much bigger protein (~650aa), finished with no issues, so I don't think that size of the protein is causing the problem.

What might be the cause of these long waiting times? Is it possible that there was an error on our side?

Best regards,
Igor

Do the ModRefiner and/or FG-MD servers take protein-bound hetreo-atoms (organic ligand, metals, ions) into account in the refinement process? If not, is there another tool that does?

Hello!
Some users before me encountered this problem of missing Beta sheets in PDB files from I-Tasser.
What I understood is that I-tasser generates PDBs without secondary structure information. The GIF in the result tab is the result of the JMol "calculate structure rama" algorithm on the PDB without secondary structure ( https://zhanglab.ccmb.med.umich.edu/bbs/?q=node/4096 ) .
PyMol and Chimera (as many others) have default algorithm to detect secondary structure (dssp , ksdssp ) , this is why the PDBs look different in JMol and PyMol/Chimera.
Anyway, I followed the instructions on the previous link to have an exact copy of the GIF showed by I-Tasser in Jmol. They work!
My big problem now is that I can't generate a PDB from Jmol that contains the secondary structures information defined by "calculate structure rama" script.
I tried many of these commands ( http://wiki.jmol.org/index.php/File_formats/Export ), but they don't work, I can just save the initial PDB without secondary structures information. (I can save these secondary structure changes only in Jmol current status)
So, do you know how can I write this pdb containing the secondary structure information derived by "calculate structure rama"?
I really hope that you will answer my question, I am sure that I could be useful for many other users too!
Cheers,
Matteo

Hi all,

I have been running the I-TASSER standalone package on my Desktop with runstyle gnuparallel multithreading on a 64-bit Linux OS. Recently, I repeated one of the jobs I submitted to the server itself and noticed structural and GO differences between the results. I have had some occasional errors that I looked up on these message boards and which were deemed as unimportant for structure accuracy. However, I was wondering if this has something to do with not updating the ITLIB (itasser library). See the script below. This is after running COACH (step 7) the first time and second time. What do we use as the PDB list when updating? Does the list in the GO folder suffice? Or do we need to make another one and put it in the PDB folder (I used download_lib.pl). Additionally, sometimes I noticed the threading was getting killed so I edited the program to use only four cores instead of the 8 that I have available. What is the issue here?

1st time:

7 run COACH to predict function: Ligand-binding site,EC number,GO terms...
/home/rrahman/I-TASSER5.1/I-TASSERmod/runCOACH.pl -pkgdir /home/rrahman/I-TASSER5.1 -libdir /home/rrahman/ITLIB -runstyle gnuparallel -protname DDB_G0271656 -model model1.pdb -datadir /home/rrahman/I-TASSER5.1/DDB_G0271656 -homoflag real -idcut 1 -LBS true -EC true -GO true
Useless use of /d modifier in transliteration operator at /tmp/root/JSD_qury/parse_blast.pl line 314.
Use of uninitialized value in printf at /home/rrahman/I-TASSER5.1/DDB_G0271656/model1/cofactor/CF_DDB_G0271656_model1.pl line 1337.
Use of uninitialized value in printf at /home/rrahman/I-TASSER5.1/DDB_G0271656/model1/cofactor/CF_DDB_G0271656_model1.pl line 1337.
Useless use of /d modifier in transliteration operator at ./parse_blast.pl line 314.
4075.22user 1190.73system 44:11.46elapsed 198%CPU (0avgtext+0avgdata 3513468maxresident)k
32990304inputs+808888outputs (284major+84600560minor)pagefaults 0swaps

2nd time:

Hi all,

I am conducting an analysis of protein sequence-structure relationships. I performed an all against all structural alignment using TMalign and would like to know if the sequence identities output by TMalign are equivalent or comparable to those from Blast or any other sequence alignment tool. I would like to use these sequence identities to generate RMSD-Sequence identity plots.

Thank you.

Hi Team,

I have submitted a job on 29th May, with Job-id S545769. But I have not received the access key on my registered mail, also I have forgotten the same.
Kindly provide me the same as its of urgent importance. Thanks

Hello, I'm writing a thesis about models predicted by I-TASSER and the corresponding energy caculated by Quantum Chemistry method. I have several questions about the Number of decoys and cluster density.
According to SPICKER, the rank of number of decoys and cluster density should be the same for the five models, as the REMC method indicated.
However, the result is as follows:

Name C-score Exp.TM-Score Exp.RMSD No.of decoys Cluster density

Model1: -2.42 0.43+-0.14 7.4+-4.2 5028 0.1417
Model2: -2.36 3419 0.1509
Model3: -3.07 2371 0.0739
Model4: -4.08 606 0.0271
Model5: -2.91 1772 0.0875
If ranked by number of decoys, then model5 should rank the fourth and model1 as the first. However, while ranked by cluster density, model 2 should rank the first. I really got confused by the ranking. And is there a formula to calculate the cluster density? I didn't find any in the paper about SPICKER.

I took the structure from the PDB database and predicted the structure by removing the homologues. So I also compare the TM-score, RMSD as well as the potential energy relative to the native structure.
The result showed that model2 is the closest to the native structure. In this case, it should be ranked as model1 instead of 2.
Looking forward to your reply & plenty of thanks.

I have run an I-TASSER 5.1 job on my computer cluster and I can't seem to find the "index.html" and "result.tar.bz2" files in my recent runs but I was able to see those files in my previous runs. Would you please enlighten me on that?

Thank you very much for your help!

Hi,

I ran I-TASSER with a contact restraint file where contact were described as CONTACT 6 30 . My file has 61 contacts in it. However, when the script make restraints with the mkres_seqname, the output is:

Hostname: Sibia
Started : Mon 25 May 2020 02:09:43 PM EDT

Path : /tmp/vincent/ITtest_force

RESTRAINT type: 1
No of Distance Restraints : 0
No of Contact Restraints : 0
Long Distance Type 1 : 0
Short Distance Type 1 : 0
Contacts TYpe 1 : 0
ending time: Mon 25 May 2020 02:09:43 PM EDT
Is it normal to have this kind of output? I looked at the predicted structure and it wasn't really different from the prediction without contact. Is I-TASSER really using restraints in my case? It would be appreciated if you could look at my results (attached).
Also, I wanted to ask if it was normal to have the 3 files (explong.dat, expshort.dat and Scontact.dat) filled with a unique 0. Are they meant to keep track of the total of each type of restraint? I tried to look up to the perl script but as I am not familiar with the way perl worked, it is difficult for me to find exactly what those files are.

The prediction is run in a script where every parameter is kept in a file called config.txt. The command is
~/I-TASSER5.1/I-TASSERmod/runI-TASSER.pl -libdir "$libdir" -datadir "$datadir" -seqname "$1" -runstyle "$runstyle" -light "$light" -hours "$hours" -homoflag "$homoflag" -idcut "$idcut" -ntemp "$ntemp" -nmodel "$nmodel" -traj "$traj" -LBS "$LBS" -EC "$EC" -GO "$GO" -restraint1 "$2" >>$1.log 2>>ITerror.log

Thanks

Vincent