possible error in make_fragments.pl and fix

Member Site Forums Rosetta 3 Rosetta 3 – General possible error in make_fragments.pl and fix

  • This topic has 2 replies, 1 voice, and was last updated 12 years ago by Anonymous.
Viewing 2 reply threads
  • Author
    Posts
    • #1256
      Anonymous

        Dear fellows,

        thank you for the newly rewritten version of make_fragments.pl in Rosetta 3.4 … it came just in time when I was struggling with the old one! In my hands/case though, there was a problem with the inclusion of secondary structure prediction from pre-calculated files. (It worked well when files were calculated on the fly with psipred / SAM installed.)

        I’ll try to be exact with my description of the errors in the current version:
        1) Filenames like “test1_psipred.ss2” when given with -psipredfile are read into $options{psipredfile}, but in the subroutine file_overrides_option is called with $options{psipred_file}. Therefore, make_fragments.pl in its current version always tries to run psipred (or SAM/porter).

        2) Inclusion of the secondary structure predictions in the flags handed to picker and in the score.cfg and quota.def files are governed by the $options{psipred/sam/porter}, which is 0 if the programs are not run, but only the .ss2 file is included with the according option, like -psipredfile *.ss2. Therefore, only if make_fragments.pl tries to run psipred/SAM/porter the results are included in the scoring.

        I came up with a fix for both making minimal changes to make_fragments.pl which I tested and which worked in my hands. A diff output file shows only the changes I made. Both files are attached as text files.
        I’d be happy about a further comment and hope everything works nicely now :)

        Greetings,
        Daniel

      • #7022
        Anonymous

          posting here with file uploading & preview didn’t work…
          Could somebody delete the other “spam” posts.. sorry!

          I’ll try to attach the diff file here.

          Daniel

        • #7023
          Anonymous

            …and the script :)

        Viewing 2 reply threads
        • You must be logged in to reply to this topic.