Command Line Interface

The annotation of PNPs require the processing of three different datasets: fragments, natural and synthetic. To better organize and orchestrate the work, snakemake workflows were developped.

Each workflow is based on a set of Python or Bash scripts, that can be used individually as well in fragments, natural and synthetic workflows.

Current scripts are regrouped in different categories and described in the table below:

List of commands

Category	Name	Description
Core	fc_classify	Classify fragment combinations by pairwise comparison of fragment hits for each molecule.
Core	fcg_annotate_pnp	Annotate FCGs with PNP information.
Core	fcg_generate	Generate fragment combination graphs using all fragment combinations found for each molecules.
Core	frags_annotate_fcp	Annotate the fragments with Fragment Combination Points (FCP).
Core	mols_dedupl	Filter out duplicate molecules and (if provided) update a reference file for processing in parallel.
Core	mols_depict	Generate 2D-coordinates for molecules.
Core	mols_draw	Draw molecules into individual files. Useful for exporting fragment images for FCG representation.
Core	mols_extract_murcko	Extract the Murcko scaffold from molecules.
Core	mols_load	Load molecules into RDKit format.
Core	mols_standardize	Apply a serie of standardization operations on molecules.
Core	mols_subset	Filter out molecules from a molecular file using another molecular file as a reference.
Miscelleneous	fhits_filter	Filter out one or several fragment hits from fragment- search and combination results using their fragment ID.
Report	report/report_fcg_chunk	Preprocess the report of the FCG information for a given chunk (optimization for cluster).
Report	report/report_fcg_concat	Concatenate all chunks of FCG informations produced by report_fcg_chunk and generate the report.
Report	report/report_fcg2	Report the number of fragment hits and fragment combinations and fragment combination graphs.
Report	report/report_fcp	Report the number of symmetry centers found in fragments.
Report	report/report_mols_count	Report the number of molecules found for a given chunk in all steps of a workflow.
Report	report/report_pnp	Report the number of PNP and NP-like molecules.
Report	report/report_prep	Report the number of passed filtered and error molecules during preparation.
Report	report/report_subset	Report the number of molecules filtered out of the input dataset during subsetting.
Report	report/report_time	Report the computation duration of a given chunk in all steps of a workflow.
Workflow management	chunk_check	Check if there are missing chunks in the output directory.
Workflow management	chunk_sdf	Split a SDF into chunks without parsing molecules.
Workflow management	concat_sdf	Concatenate SDF chunks into a single SDF without parsing molecules.
Workflow management	concat_synonyms	Concatenate all synonym files (references for identifying duplicate molecules) into a single file.
Workflow management	mols_concat	Concatenate two molecular files.
Workflow management	mols_count	Count the number of molecules in a molecular file.
Workflow management	report_protocol	DEPRECATED. Reports are now generated automatically during workflow executions.
Workflow management	run_protocol_fc	Run a snakemake workflow (either fragments or natural or synthetic).

Some commands are used in the workflows (Core and Report categories), whereas some others were used only once but were kept in the tool for convenience. One example would be the the fhits_filter script, which was used to filter out benzene fhits from the fragment search and fragment combination results, without having to restart these computations.