Figure S1 - SWEETLEAD: an In Silico Database of Approved Drugs, Regulated Chemicals, and Herbal Isolates for Computer-Aided Drug Discovery

Example Consensus Building for Apomorphine Hydrochloride. Step 1) To begin the consensus building process, each Reference Database is queried by the compound name, and Structural Database IDs are collected. The resulting IDs are tallied by assigning each Structural Database ID a vote equal to 1/(# Structural Database IDs returned by that Reference Database). So, in this example, PubChem ID 107882 is the most popular PubChem ID by obtaining 1.25 votes (1 for being the only result returned by PubChem, and .25 for being one of 4 IDs returned by ChemSpider). Step 2) The structure associated with each winning Structural Database ID is obtained, and standardized according to the text. The structures are then associated with all Structural Database IDs which return the structure after standardization (various salt forms, protonation states, etc.) Step 3) The most popular structure is identified. In this case, there is a consensus structure as all Reference Databases pointed towards Structural Database IDs matching the most popular structure. Step 4) The winning structure is associated with the drug name, winning Structural Database IDs, synonyms associated with the winning Structural Database IDs, and entered into the SWEETLEAD database. (PDF)