Hi Amy,
Ok, this is really confusing.
file phremd_1.cpout
returns
phremd_1.cpout: ASCII text
even though
grep “Solvent pH” phremd_1.cpout
returns
Binary file phremd_1.cpout matches
Maybe the file is a mix of ASCII (what I see when I open it) and binary (unprintable characters).
No, the file was never tarred.
I had to put other simulations ahead of re-doing this one, but I’ll see if I can reproduce the issue. I know that other pH-REMD simulations on the same workstation I’ve run do not have this problem (fingers crossed on the one running right now.)
Thanks again for your help. I’ll update this thread on the reproducibility point when I have the results for this specific simulation.
Best,
Matthew
> On Aug 19, 2024, at 5:46 PM, He, Amy <he.1768.buckeyemail.osu.edu> wrote:
>
> You don't often get email from he.1768.buckeyemail.osu.edu. Learn why this is important
> Hi Matthew, I’m not sure why the output appears to be a binary file. What does it say when you do “file phremd_1.cpout”? Was it a .tar file, by any chance? This is the only possibility I can think of, given the fact that you were able to recover some texts by tr and grep. But I can’t explain the irregularities. If you can reproduce this, it might be something worth reporting to the develop team.
> Best regards,
> Amy
> From: Guberman-Pfeffer, Matthew <Matthew_Guberman-Pfe.baylor.edu>
> Date: Sunday, August 18, 2024 at 9:17 PM
> To: He, Amy <he.1768.buckeyemail.osu.edu>
> Cc: AMBER Mailing List <amber.ambermd.org>
> Subject: Re: Unclear cphstats warning
> Hi Amy! Thanks so much for the suggestion. Something went wrong in the writing of the cpout files. Here’s what I found: I went to grep “Solvent pH” and got: Binary file 001/phremd_1. cpout matches This is true for all four replicates for thisHi Amy!
> Thanks so much for the suggestion. Something went wrong in the writing of the cpout files. Here’s what I found:
> I went to grep “Solvent pH” and got:
> Binary file 001/phremd_1.cpout matches
>
> This is true for all four replicates for this particular run (indexed with “1”).
> But when I open he cpout file with vim, it looks like a normal text file.
> My next step was to analyze the pH listed for each residue. To do that, I needed to convert the binary files to text files with:
> tr -cd '\11\12\15\40-\176' < phremd_1.cpout > phremd_1_text.cpout
> Then I ran: grep -E "Residue\s+[0-9]+\s+State:\s+[0-9]+\s+pH:\s+[0-9]+\.[0-9]+" yourfile.txt | sort -k6,6n
>
> And I found that some of the lines were wrongly formatted; for example:
> Residue 9 State: 0 pH: 15.500Residue 21 State: 1 pH: 14.000
> Residue 9 State: 0 pH: 15.500RResidue 10 State: 1 pH: 15.500
> RResidue 10 State: 1 pH: 15.500
> Residue 10 State: 1 pResidue 16 State: 1 pH: 14.500
> Residue 19 State: 0 pHResidue 16 State: 1 pH: 15.000
> I find similar irregularities in the original file that grep thinks is in binary:
> Residue Residue 40 State: 0 pHResidue 2 State: 1 pH: 15.500
> It looks to me like something went wrong with the synchronization where multiple processes were writing simultaneously to the file. What do you think?
> What troubles me is: Why? I’ve ran dozens of pH-REMD simulations on the same computer and this is the only instance of this problem.
> I think I’m going to re-run this simulation. But if there is someway I can know how to prevent whatever happened, that would be great because the simulation takes about a day.
> Best,
> Matthew
>
>
> On Aug 18, 2024, at 11:28 PM, He, Amy <he.1768.buckeyemail.osu.edu> wrote:
>
> Hi Matthew! I might be wrong, but I think this warning message often appears when there’s a problem with the first line in the cpout files, “Solvent pH”, which is, when the data in your cpout files contain a pH that’s not reported in the first line of any of the provided cpout files. If you’re very sure that the data in your cpout files are correct, you can re-distribute the data by your own codes without looking at Solvent pH at the beginning.
> Best regards,
> --
> Amy He
> Forli Lab . TSRI
> yihe.scripps.edu
> From: Guberman-Pfeffer, Matthew via AMBER <amber.ambermd.org>
> Date: Sunday, August 18, 2024 at 12:02 PM
> To: AMBER Mailing List <amber.ambermd.org>
> Subject: [AMBER] Unclear cphstats warning
>
>
> Dear AMBER community,
>
> I ran a 4-way pH-REMD in two chunks (indexed as 0 and 1). When processing the output form the second of these chunks with the following command:
> cphstats --fix-remd reordered_1 001/phremd_1.cpout 002/phremd_1.cpout 003/phremd_1.cpout 004/phremd_1.cpout
>
> I get:
> WARNING: not all data records were assigned to a specific cpout file
> this can happen, for instance, if you did *not* provide all
> of the replica cpout files to cphstats to process. Check
> your results carefully!
>
> But I did provide the cpout file for each of the four replicas. Something else is presumably wrong. How can I figure out what is causing the warning? I’ve run the same command in other directories for differ 4-way replicas and many more chunks and get no warning. Something seems to have happened in this particular case.
>
> Best,
> Matthew
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> https://urldefense.com/v3/__http://lists.ambermd.org/mailman/listinfo/amber__;!!KGKeukY!2mHqf4s1EGxhUbnMdvgWrUroyEwc4TvQW3kbNM9f6WUqXbn_pu_fWG0u8EMC4Ie-S_WnjCDPa1jo-OSJWEh9XU2h9j0$
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Aug 19 2024 - 15:00:03 PDT
