Files in Data directory

Individual files

• Descriptor configuration files cfp-7-1.xml, ecfp-4.xml are based on samples provided in JChem distribution (in jchem/examples/config/).

• Text file floatdesc.txt contains an example 2D float vector descriptor in line format <ID> <X> <Y>.

• Text file floats-1d.txt contains an example 1D float vector descriptor in line format <ID> <X>. This file contains empty lines and comment lines starting with character #. The textual <ID> field reflects the actual value <X> for each item.

• Details on script sanitize-prof.js can be found in document Profiling and execution statistics.

Frontend customization examples (jsclient-additional/)

For details see document Using MadFast Web UI JS library (experimental).

Molecule sets (molecules/)

Details on these and further publicly available molecule sets can be found in document Prepare example molecule sets.

• File molecules/vitamins/vitamins.smi contains example molecules from http://en.wikipedia.org/wiki/Vitamins.

• File molecules/antibiotics/antibiotics.smi contains list of antibiotics based on https://en.wikipedia.org/wiki/List_of_antibiotics table "By class".

• File molecules/who-essential-medicines/who-essential-medicines.smi contains structures from WHO Model List of Essential Medicines (*adult list* of 19th edition, April 2015), based on https://en.wikipedia.org/wiki/WHO_Model_List_of_Essential_Medicines.

• File molecules/vitamins/vitamins-mod.smi contains modified versions of molecules in file vitamins.smi. A single carbon atom ("C") is added to the beginning of the SMILES strings in the source file (this results in valid but slightly different SMILES sources). String fragment "- MODIFIED" is also appended to the names. This file can be used to benchmark search against larger public databases.

• Directory molecules/drugbank/ contains the DrugBank Open Data dataset, available from http://www.drugbank.ca/releases/latest#open-data under Creative Common’s CC0 International License.

• Directory molecules/nci/ contains the NCI Release 1 dataset, available from http://cactus.nci.nih.gov/download/nci/. According to the publisher of this dataset the structures are in the public domain.

• Directory molecules/chembl/ contains ChEMBL compounds from http://www.ebi.ac.uk/chembl. The version of ChEMBL is chembl_21. For details see file molecules/chembl/README.html. Please note that structure CHEMBL2146209 (https://www.ebi.ac.uk/chembl/compound/inspect/CHEMBL2146209) is omitted from the SMILES version.

• Directory molecules/pubchem-compound/ contains 1k 10k and 100k random subsets from the PubChem Compound database.