I would like to use your program Probis for a binding site comparison. I followed the steps in the Users manual. However, when running the step
../probis -ncpu 8 -surfdb -local -sfile srfs.txt -f1 1phrA.srf -c1 A -nosql example.nosql
and I optain not example.nosql file. I tried both the precompiled 64bit linux version and the source code compiled on my machine. Both give the same error. I would be grateful for any help.
Dear Nadine,
The command should be on one line like this:
../probis -ncpu 8 -surfdb -local -sfile srfs.txt -f1 1phrA.srf -c1 A -nosql example.nosql
Your example streatches two lines.
If this is not the error you were looking for, then could you please send me zipped directory with all files?
Hi Janez,
I am aware that the command needs to be in one line. Probably the E-Mail program broke into two parts.
/local_software/probis_source/probis/probis -surfdb -local -sfile srfs.txt -f1 ../probis_bsite_files_30/1j04_LLP_1000_A.srf -c1 A -nosql 1j04_LLP_1000_A.nosql
I have attached a sample dataset. The commands are generated by a python script in the folder probis_analyis_30.
I would be highly grateful for your help with this.
Hi Nadine,
The solution is to add chain_id's as the second column in the srfs.txt file. Attached is modified srfs.txt file from your example that works.
Hi Janez,
with your suggested modification I am now able to generate nosql-files successfully, thanks. However, in these nosql files there are only data for some alignments shown. I would like to do an all against all comparison. For this I need alignments even if they are bad, for all protein binding sites. Is there a cutoff value I can set differently to prevent that only some alignments are not shown/generated?
Use this option:
-param PAR_FILE Read parameters from the specified parameter file.
Hi Janez,
thank you for your help. However, unfortunately the output did not change with the param option.
probis -surfdb -local -sfile srfs.txt -param parameters.inp -f1 ../probis_bsite_files_50/1j04_LLP_1000_A.srf -c1 A -nosql 1j04_LLP_1000_A.nosql
Where the srfs.txt file contains all the srf files to compare with.
Hi Nadine,
This took me a while. Add the -noprune -nofp options to the command line!
Hi Janez,
great, that worked, thank you.
Hi Janez,
I have been happy to early. It is true for my first sample case but not for the others. You can use the files I sent you. The protein 1a8i gets comparisons with all others, which is fine. But, the other proteins have only comparisons with a fraction of the other proteins in their nosql files= .
Hi Janez,
did you have time to have a look? Can you tell if this is something general, i.e. an all against all comparison is not possible in general? Or if it is something system dependent? Thank you for your help!
Hi Nadine,
Add this line to the parameters.inp file to raise the similarity threshold:
THRMSD 1.8
In some rare cases it might still not output all lines, regard those proteins as non-similar.