Calculating Solvent Accessibility With Wesa . How To Compile And Run On Os X?

Question

[shane.neeley]

Was just hoping there was someone on this forum familiar with this program WESA.

http://pipe.scs.fsu.edu/wesa/

I want to get it running on my own computer. Mac OS X mountain lion. The software (under download software link) says it is primarily intended for linux systems. But I was wondering what I would have to change to run it.

I went through the instructions in the readme (downloading psi-blast, nr database, and changing directories in the run_WESA.pl file, and doing a 'make' of the program). But the program chokes still.

WESA $ perl run_WESA.pl test1.seq
test1 88
running neural network...
Neural Network went wrong at run_WESA.pl line 67.

The line 67 is where perl uses 'system' to run the program. This is why I think it has to do with linux vs. mac.

 print "running neural network...\n";
        !system("$exe_dir/NN_prd.x $src $wgt_dir/wgt.neu")
           or die "Neural Network went wrong";

I may be way out of my league with trying to run this on OS X.

Additionally, can you think of any other programs to predict buried or exposed residues with the source code available?

Thanks again. :) Shane

Answer 1

[João Rodrigues]

What exactly is the error message pointing at (that line 67?)

A quick google search gave this as a possible alternative: http://sable.cchmc.org/

EDIT:

When I run the example myself it gives the same error message. However, running the example manually works. It seems there is something wrong with the paths, maybe too long? To clarify:

joaorodrigues@WESA: ./run_WESA.pl test1.seq 
test1 88
running neural network...
/Users/joaorodrigues/Downloads/WESA/bin/NN_prd.x test1 /Users/joaorodrigues/Downloads/WESA/wgt/wgt.neu
Neural Network went wrong at ./run_WESA.pl line 68.

and

./bin/NN_prd.x test1 /Users/joaorodrigues/Downloads/WESA/wgt/wgt.neu
Abort trap

But this works:

joaorodrigues@WESA: ./bin/NN_prd.x test1 wgt/wgt.neu

So my guess is that the path is too long.. ? The perl script seems just a bunch of straightforward commands so maybe you can try to run them manually and see how it goes. If the problem is indeed the path, then you can just write a bash script or something to work with those paths (maybe in tmp or something).

[shane.neeley]

Edited the question. Yes SABLE is downloadable here: http://folding.cchmc.org/software/software.html But it says for linux / windows so I might have the same problems.

[João Rodrigues]

Check my EDIT above.

[shane.neeley]

Thanks. You must be a pretty good programmer to notice that. I showed this to a very experienced C programmer and he said that it was because they paths were too long. So he edited the NNprd.c file to accept longer path lengths. So then the perl program started working with the test files. Where I am stuck now is using my own files. Here is the error: "PSI-BLAST went wrong at ./runWESA.pl line 64." which where the psiblast commands for the seq file are invoked. I tried downloading older versions of blast pre-2005, but that didn't help. Are you interested in working on this with me?

[João Rodrigues]

Glad I could help a bit. I can try to help you out. What is the error exactly with the BLAST? Try running the command that the perl script invokes and see what the error message is.