Error about permission denied when using the docker version.

[otzfadia]

Hello, I am running the following:

sudo python3 run_surfmap_image.py -d OREN/ -pdb OREN/3fav.pdb -tomap binding_sites
Running: /home/surfmap/MSMS/pdb_to_xyzr /input/3fav.pdb > ./shells/3fav.xyzr-tmp ...

surface property mapping: binding_sites
Traceback (most recent call last):
  File "/home/surfmap/tools/SurfmapTools.py", line 178, in <module>
    main()
  File "/home/surfmap/tools/SurfmapTools.py", line 93, in main
    out = open(outdir + "%s_%s_partlist.out"%(pdbName.split(".")[0],tomap),"w")
PermissionError: [Errno 13] Permission denied: '/output//3fav_bfactor_partlist.out'

Mapping of the residues given in input failed. This can be due to a malformed residue file or a mistake in the reisdue numbering, type or chain.

Format should be the following:
chain residuenumber residuetype
example: A  5 LEU

Exiting now.

Why I get permission denied, even when I use sudo? Cheers! Oren

[nchenche]

Hello,

First of all, thanks for using SURFMAP (or at least trying to :) ).

docker case

Unfortunately I wasn't able to reproduce your error. It might come from running the docker with root privileges but I am not sure of that. You could try following the procedure described here to run docker without root privileges.

However, when trying to reproduce your error, I have noticed something wrong: the directory output given after the -d optional argument must be written without /, so instead of -d OREN/ you should do -d OREN (it will be corrected in a later version).

My additional advises to help you resolving the issue from the docker are:

• try the same command with a file given in example (SURFMAP/example/1gv3_A_binding_sites.pdb). If it does work, the issue could come from your input pdb file.
  • python run_surfmap.py -d OREN -pdb example/1gv3_A_binding_sites.pdb -tomap binding_sites
• try to run surfmap with your pdb file but without specifying any directory output (i.e. no -d option)

Also, if you are still facing any issue, could you please share your input pdb file ? I also need to add that using -tomap binding_sites requires the pdb file to be edited with the b-factor column containing an integer value (0, 1, 2, ...), with 0 for atoms outside of any binding site, and 1, 2 or more for atoms being present in binding site 1, binding site 2 etc (the numbers are used to label different binding site with different colors). That being said, no error should occur if the input pdb file is filled with original b-factor values, just a strange colored map.

local running case

For this case, it appears that you don't have numpy installed in your python library. I thus deduce that you might not have installed required python libraries through the following command: pip3 install -r requirements.txt (requirements.txt is at the root of the SURFMAP directory).

[otzfadia]

Thank you very much for your speedy reply! Indeed the pip3 did the trick. I now have a question regarding interpetation of the results. May I contact you directly? Cheers! Oren

[otzfadia]

Archive.zip Files I used.

[nchenche]

You're welcome and thank you for the archive, I will take a look.

You can contact us at the following mails:

• anne.lopes@i2bc.paris-saclay.fr
• hugo.schweke@weizmann.ac.il
• nicolas.chevrollier@laposte.net

Please put all of us in copy to increase the speed of your answer(s).

Cheers

[nchenche]

So I looked at your archive. From what I saw, I understand that you tried to "find" and map the binding sites of components of a multimeric protein. Unfortunately, SURMAP does not "find" binding sites, it only maps them as long as their residues have already been identified and this information been given in the b-factor column of an edited pdb file by the user.

For instance, let's say for simplicity that you know that the residues I, F, and Q of the chain A constitute a binding site. To map those residues with SURFMAP, you will need to edit the pdb file in the following way:

• enter a value of 1 in the b-factor column of each atoms constituting the residues I, F and Q from the chain A
• enter a value of 0 in the b-factor column of each atoms of residues that are not part of the binding site

Hope I didn't misunderstand your aim of using SURFMAP here and answer your question.

I will try (maybe tomorrow) to make a script to help users modify a pdb file from simplified information on the known residues of a binding site in order to avoid editing "manually" a pdb file, if that could help.

[otzfadia]

Thank you very much! Yes we di have the locations of the binding sites, so we need to edit the pdb files we use as input. Your kind suggestion for a script will be very helpful! Cheers! Oren