The need for navigating chemical space has become more important due to the increasing size and diversity of chemical biological databases (e.g., ChemSpider, DrugBank, ChEMBL, ToxCast). To do so, modelers typically rely on projection techniques applied to series of quantitative molecular descriptors directly computed from two-dimensional chemical structures. However, the multiple cheminformatics steps required to compute and visualize chemical space are technical, necessitate coding skills, and thus represent a real obstacle for non-specialists. Inspired by the popular Google Maps application, we developed the ChemMaps.com webserver to easily navigate chemical spaces. The first version of ChemMaps was developed to browse and visualize the space of 2,000 FDA-approved drugs and over 6,000 drug candidates. Each compound was initially characterized using a large set of molecular descriptors including 1D-2D RDKIT descriptors and 3D PADEL descriptors (238 1D-2D and 44 3D after removing correlated descriptors). Principal Component Analysis was used to project compounds in three-dimensional space, where compounds’ coordinates in the first two dimensions were calculated using 1D-2D descriptors, and the third dimension (Z axis) was determined using 3D descriptors only. To optimize the representation of the space and the interactive, user-friendly navigation experience, we developed the ChemMaps.com webserver using modern 3D-optimized web technologies such as HTML5, JavaScript, and CGI. The chemical coverage is now being expanded to include environmental chemical space based on the U.S. EPA TSCA inventory, as well as toxicological categorizations based on curated animal study data and predictive high-throughput screening signatures. Users accessing ChemMaps.com can immediately explore the entire compound library using a responsive, mouse-based, easy-to-use navigation tool. Since all information and coordinates are pre-computed, the browsing is instantaneous and does not require computational skills. Similar to searching Google Maps for a specific address, users can search ChemMaps via a dedicated search bar (e.g., name, indications, pharmacological class, toxicity values) and visualize the space with options to zoom in on chemical “neighborhoods”. Additional browsing, searching, and exporting options are underway, including tools to support read-across and chemical risk assessment.

Exploring chemical space using ChemMaps.com

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.