ChemHTPS Modules and the Config File

ChemHTPS is called through a simple shell command along with arguments signaling the modules that will be triggered. The modules are:

  • library generator
  • job generator
  • job feeder

The config file should closely follow the template of the initial config file that is created once the project is set-up. There are general as well as module-specific options as follows:

  • log-file: Specify the log-file for a call of chemhtpsshell.

    • Expected type: string file-name
    • Default value: “project_name”.log

  • error-file: Specify the error-file for a call of chemhtpsshell.

    • Expected type: string file-name
    • Default value: “project_name”.err

library generator module:

-called by –generatelib parser argument
  • generation_file: Specify the ChemLG config-file to be used for this module call for ChemLG execution.

    • Expected type: string file-name from current location
    • Default value: None

  • building_block: Specify the ChemLG building-block file to be used for this module call for ChemLG execution.

    • Expected type: string file-name from current location
    • Default value: None

  • cores: Specify the number of cpus to be used for this run of library generator or the ChemLG call (integers).

    • Expected type: integer
    • Default value: None

  • clusters: cluster or partitions where ChemLG will be executed during the library generator module.

    • Expected type: string cluster/partition, whichever applicable or “run locally”
    • Default value: None

job generator module:

-called by –generatejobs parser argument

We have made a separate guide as to how an input file is created from a template.

  • library_in: Specify the library of molecules to run jobs for.

    • Expected type: string library-name from within ./screeninglib folder
    • Default value: None

  • job_type: Specify the class of jobs regarding their resource-expenses.

    • Expected type: “short” or “long” strings, depending on whether the jobs will take short or long times
    • Default value: None

  • program: Specify the program that jobs will be run for and kind of template-files that will be called.

    • Expected type: “QCHEM” or “ORCA”, string
    • Default value: None

  • template: Specify the template file that we used to generate job-files from.

    • Expected type: string, the file location within “./job_templates/$program/” folder
    • Default value: None

job feeder module:

  • called by –feedjobs parser argument
  • feed_local: Specify whether the jobs will be submitted remotely or from within the local node.

    • Expected type: string, “TRUE” or “FALSE”
    • Default value: None

  • job_sched: Specify the job-scheduler to be used.

    • Expected type: string
    • Default value: “SLURM”

Along with this guide about how the config-file options operate, please do check out a short tutorial we have compiled at: