The following python script and library are needed to automatically submit DISPLAR jobs for a large number of proteins and obtain predictions from your own computer.
Usage
(1) Save these 2 files in a directory on your computer. From there, to submit a DISPLAR job, type the command:
| python DISPLAR-submit.py $1 $2 $3 $4 > $1.html |
where
- $1 = submission name. (It should be unique for each DISPLAR job; otherwise prediction will be overwritten.)
- $2 = your email address
- $3 = the protein chain IDs in the PDB file on which you want to predict a DNA-binding site. $3 has the format "_" or "A" or "A,B,C" and must match the chains in the PDB file.
- $4 = PDB filename. (You can specify the full path of the file.)
- $1.html contains information on the job submitted and a link to the prediction. To view it, simply type:
(2) Wait for the job to finish (typically a few minutes); the prediction is then sent to you by e-mail. The prediction is also displayed when you click the link in $1.html. You can save the prediction to a file, e.g., $1.out, by typing:
| wget -O $1.out -F -i $1.html |
(3) Combining the above two steps allows you to automatically submit multiple jobs. Copy the following lines to a file called displar-submit.scr:
| python DISPLAR-submit.py $1 $2 $3 $4 > $1.html |
| sleep 900 |
| wget -O $1.out -F -i $1.html |
Note: "sleep 900" is added to allow a wait period of 900 seconds before collecting prediction and submitting a new job.
Change the mode of displar-submit.scr to executable (chmod +x displar-submit.scr), and create a run file (e.g., displar-submit.run) that contains one line for each protein. The run file looks like this:
| ./displar-submit.scr job1 email chain1 pdb1 |
| ./displar-submit.scr job2 email chain2 pdb2 |
etc. Change the mode of the run file to executable and then type the name of the run file. The predictions will be saved in job1.out, job2,out, etc.
https://pipe.rcc.fsu.edu
|