luancarvalhomartins/PyAutoFEP

Invalid atom number 3804 in indexfile

Opened this issue · 10 comments

whe280 commented

Hi,

an error occured from */tutorial/FXR_12-FXR_74/water/lambda0/nvt/nvt_grompp.log, but index.ndx and nve.gro have 3803 atoms:
`
atom name 19 in FXR_12-FXR_74/water/lambda0/nvt/../fullsystem_161000_17122023.top and FXR_12-FXR_74/water/lambda0/nve/nve.gro does not match (O0I_const_idx19 - O0I_c)

atom name 20 in FXR_12-FXR_74/water/lambda0/nvt/../fullsystem_161000_17122023.top and FXR_12-FXR_74/water/lambda0/nve/nve.gro does not match (O0J_const_idx20 - O0J_c)

(more than 20 non-matching atom names)

WARNING 1 [file fullsystem_161000_17122023.top, line 55]:
58 non-matching atom names
atom names from
FXR_12-FXR_74/water/lambda0/nvt/../fullsystem_161000_17122023.top will be
used
atom names from FXR_12-FXR_74/water/lambda0/nve/nve.gro will be ignored

Program: gmx grompp, version 2021.6
Source file: src/gromacs/gmxpreprocess/readir.cpp (line 2815)

Fatal error:
Invalid atom number 3804 in indexfile

For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors
`

Will someone please advise on how I can resolve this?
Thanks!

whe280 commented

The relevant documents are attached.
water.zip

AzraelXu commented

The issue arises in the latest version of PYAUTOFEP code. You can try using a previous version of PYAUTOFEP or opt to use a pre-established system. As for the problem with the NVT simulation of the water leg in the latest version, it will require the author @luancarvalhomartins to address and resolve it.

whe280 commented

The issue arises in the latest version of PYAUTOFEP code. You can try using a previous version of PYAUTOFEP or opt to use a pre-established system. As for the problem with the NVT simulation of the water leg in the latest version, it will require the author @luancarvalhomartins to address and resolve it.

Thanks a lot！I have successfully run PyAutoFEP using a pre-established system, and I will try using automatic system builder with previous version of PyAutoFEP.

pcardonap commented

The issue arises in the latest version of PYAUTOFEP code. You can try using a previous version of PYAUTOFEP or opt to use a pre-established system. As for the problem with the NVT simulation of the water leg in the latest version, it will require the author @luancarvalhomartins to address and resolve it.

Would you mind telling me from which commit onward is this issue occurring? @AzraelXu

cmwoodley commented

I've had the same problem and also an issue with the water index file being referenced in the wrong location. I think this issue might be due to the index file generated from the solvated and neutralised systems not being added from the temporary directory to the .bin file.

I'm using this bash script to modify the .sh files as a workaround, which seems to work:

for folder in ./LIG*
do
    cd $folder
    
    for file in ./*.sh
    do
        sed -i -e 's@/tmp/tmp[^/]*/PMX_FEP/@@g' "$file"
        done
    
    prot_build=$(find ./protein/ -type d -name 'build_system*' -exec basename {} \;)
    wat_build=$(find ./water/ -type d -name 'build_water*' -exec basename {} \;)
    
    sed -i -e "s@protein/${prot_build}@water/${wat_build}@g" runall*water.sh
    
    cd protein/build_system*
    
    gmx_mpi make_ndx -f fullsystem_step8*.pdb -o index.ndx <<EOF
    1 | 13
    q
EOF
    
    cd ../../../
done

My ligands are name LIG_0, LIG_1 etc.

AzraelXu commented

The issue arises in the latest version of PYAUTOFEP code. You can try using a previous version of PYAUTOFEP or opt to use a pre-established system. As for the problem with the NVT simulation of the water leg in the latest version, it will require the author @luancarvalhomartins to address and resolve it.

Thanks a lot！I have successfully run PyAutoFEP using a pre-established system, and I will try using automatic system builder with previous version of PyAutoFEP.

You're welcome! I have a question for you as well—does the pre-established system you're using include ligands? I attempted to run FEP with a pre-established system containing water and ions, but encountered failure with an error message indicating a missing ligand in the index. Do you happen to know what could be causing this issue?
image

whe280 commented
whe280 commented
AzraelXu commented

My pre-established system doesn't contain ligands. I've encountered the error , and because I haven't worked on the project for so long, I can't remember the details. It seems to be related to the mdanalysis package version or setting selection_method=mdanalysis or checking if the water model is appropriate. | Subject | Re: [luancarvalhomartins/PyAutoFEP] Invalid atom number 3804 in indexfile (Issue #131) | The issue arises in the latest version of PYAUTOFEP code. You can try using a previous version of PYAUTOFEP or opt to use a pre-established system. As for the problem with the NVT simulation of the water leg in the latest version, it will require the author @luancarvalhomartins to address and resolve it. Thanks a lot！I have successfully run PyAutoFEP using a pre-established system, and I will try using automatic system builder with previous version of PyAutoFEP. You're welcome! I have a question for you as well—does the pre-established system you're using include ligands? I attempted to run FEP with a pre-established system containing water and ions, but encountered failure with an error message indicating a missing ligand in the index. Do you happen to know what could be causing this issue? image.png (view on web) — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

Hi Wen, it seems that selection_method=mdanalysis works ,thank you very much.
By the way , could you please inform us of your place of employment?

elapsebos commented

I've had the same problem and also an issue with the water index file being referenced in the wrong location. I think this issue might be due to the index file generated from the solvated and neutralised systems not being added from the temporary directory to the .bin file.

I'm using this bash script to modify the .sh files as a workaround, which seems to work:

for folder in ./LIG*
do
    cd $folder
    
    for file in ./*.sh
    do
        sed -i -e 's@/tmp/tmp[^/]*/PMX_FEP/@@g' "$file"
        done
    
    prot_build=$(find ./protein/ -type d -name 'build_system*' -exec basename {} \;)
    wat_build=$(find ./water/ -type d -name 'build_water*' -exec basename {} \;)
    
    sed -i -e "s@protein/${prot_build}@water/${wat_build}@g" runall*water.sh
    
    cd protein/build_system*
    
    gmx_mpi make_ndx -f fullsystem_step8*.pdb -o index.ndx <<EOF
    1 | 13
    q
EOF
    
    cd ../../../
done

My ligands are name LIG_0, LIG_1 etc.

Hello, can you tell me why there is a regeneration for index.ndx as the last step, I think the sed command for all runall*water.sh files is a simply solution to bypass the problem in indexfile as the poster said. Are there any issue else for further running I do not noticed before?
Thanks :)