Recurring pLink error

[eric.merkley]

Hi,
My plink output gives the following errors:

"This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.

terminate called after throwing an instance of 'std::runtime_error'

  what():  std::bad_alloc      at XLinkPreProc::Run() in the function CPreProcess::Run.

      at CXLinkOpenFlow::_Preproc() in the function CPreProcess::Run.

      at CXLinkOpenFlow::Run() in the function CXLinkOpenFlow::_PreProc

      at SearchEngine.cpp::_SearchEngine_Callback() in the function CFlow::Run"

and
"Failed to run normal search: [Error 2] The system cannot find the file specified: 'C:\\Users\\merk000\\Documents\\ForJonJacobs\\UBL_pLink_01\\Mock_12hr_Bot_01_Run1\\1.sample\\search\\Mock11_qry.proteins.txt'"

I am running an open-mode search with a genome-sized (mouse) database and a custom linker.  pLink.ini and xlink.ini are attached.  In addition to these errors, the 2.results folder is never created and, as the second error states, some files in 1.sample are also never written.

I think this is related to a fasta file size problem.  When I truncate the file to the first 50 proteins, the search completes without errors and all expected files are output.  But when I truncate the mgf file, the search also completes, even with the large fasta file.

I asked about this once before and someone suggested that it had to do with naming the sequence file .txt instead of .fasta, but I have not done that.  It also does not seem to be an issue with the mgf file, as other posts on this forum suggested, because the truncated file works fine.  The only non-alphnumeric character in the mgf file TITLE lines is "."

Does anyone have experience with large fasta files that could give me a pointer?

Thanks,
-Eric

[Alice Zhao]

I'm wondering if we have the same issue. I was getting the terminate/runtime error and the same subsequent errors because pLink was generating a huge temporary file and it occupied the entire drive, and could not write any more. Have you checked if this was the case for you?

[Shengbo Fan]

Hi,

This problem seems to be the memory problem. The std::bad_alloc error means that system can't allocate enough memory for the search program. This sometimes happens when there is a large database with a large number of spectra, and for 32-bit program on Windows, at most 2G memory are allowed. 

Make sure that the computer is not running other large memory-cost program, and if that still doesn't work, the truncation is needed. May either truncate the database or the spectra, search multiple times, and at last merge them.

[eric.merkley]

How would the truncation of either the mgf file or the fasta file influence the FDR calculation?  Can you still trust that the FDR of the merged results is the same as that reported by pLink for the separated parts?

Thanks,
Eric

[Shengbo Fan]

pLink provides functions for that situation. For example,

Total.mgf

Split it to 

1.mgf, 2.mgf, 3.mgf, ...

Put these mgf files into one folder, like,

Total

And specify

sample1.spectra.path=...\Total\

The program will search these mgf files one by one, and at last merge them together. The FDR will be calculated after the merge, so that will be the same as they are in one file. It's not like the way that the FDR is calculated separately, and merge the results after FDR control together, which will surely have some influence of the total FDR.

[Alice Zhao]

Can you provide some guidance on how to split one mgf into multiple mgf files?

[eric.merkley]

Alice,

I found a perl script that is part of the perl module InSilicoSpectra.  It's called split-mgf.pl.  Download the zipped tar archive, unzip it, unpack it, and look in the scripts directory.  No need to install anything (assuming you already have perl).  I'm runnin Cygwin under windows 7 and it worked right out of the box.

Here's the link:  https://metacpan.org/pod/distribution/InSilicoSpectro/scripts/split-mgf.pl

-Eric