FAQ


Q: Can ARCIMBOLDO programs run on a single workstation or laptop?
A: Yes, ARCIMBOLDO_LITE, ARCIMBOLDO_BORGES and ARCIMBOLDO_SHREDDER can run on a single computer in standalone mode.

Q: Do Phaser and SHELXE have to be installed to be accessible on every node in the grid?
A: Within SGE it is necessary. Even if condor supports the transfer of executable files, PHASER is dynamically linked, and requires external libraries to run and the SHELXE binaries are heavy to transfer every time. It is preferable to have a local copy on each machine or an accessible installation available through NFS.

Q: How do I choose a library as ARCIMBOLDO_BORGES hypothesis?
A: A template from a library can be based knowledge about the protein and what kind of fold is expected. If not at all known, from a secondary structure prediction one can derive secondary structure and try all the simple possibilities, ie. parallel / antiparallel helices or parallel, antiparallel, parallel-antiparallel beta strands libraries.

Q: Can I reuse BORGES libraries?
A: For tipical local fold with broad defaults, the same library may be suited for many cases. Some typical libraries can be downloaded with the distribution.

Q: Can I use my own library in ARCIMBOLDO_BORGES?
A: Yes, provided that it fulfills the following:

  • Models are superimposed
  • B-factor are set to the same value (25.00) unless for reasonable knowledge that the context from which they are extracted is similar to the one in the unknown structure.
  • Name of the models must follow the convention PDBi_n_n.pdb (e.g. 4ETO_0_0.pdb)
  • Variable size in the models will difficult comparison between LLGs, and we have always work with same size models, but ARCIMBOLDO-BORGES will accept it as input.


Q: Do your programs run with any PHASER/SHELXE version?
A: They surely work with latest SHELXE (2016 version) and PHASER versions 2.7.x.

Q: Do your programs support parallel computing using SMP/MPI technology?
A: No, we do not support multi-core jobs using SMP or MPI technology. Our programs internally distribute PHASER and SHELXE processes over a Grid with a Condor, Sun Grid Engine or Torque middleware or parallelized as independent process (one per core) in a single workstation.

Q: Is it possible to use ARCIMBOLDO_LITE with non-peptide models?
A: Yes. Different types of search fragments may be used. For example, standard compounds like Heme groups or iron-sulfur clusters. However, considering that the autotracing feature in SHELXE only traces main chain of polypetides, any other type of fragment (like DNA, for example) is only useful to obtain an initial map in which to trace the protein atoms and get a more complete structure.

Q: Do your programs run using the hyperthreading cores?
A: No, by default, only physical cores are claimed. Therefore, if your system has 16 physical cores, 15 PHASER or SHELXE jobs will be run in parallel, while one core is saved for the main program to run. However the program may use all hyperthreading cores with keyword force_core: Number_of_cores.

Q: Successful execution has been tested on the following distributions:
A: Successful execution has been tested on the following distributions:

  • Mavericks
  • Yosemite
  • El Capitan
  • Sierra