Through extraction of bioactivity data from peer-reviewed medicinal chemistry literature, the ChEMBL database has become one of the largest sources of open chemogenomic data. However, high quality small molecules are currently available for only ~10% of the proteins coded by the human genome. In attempting to expand the druggable proteome as part of the Illuminating the Druggable Genome project, patents were considered as a source of additional data. Novel compounds are often published in patent documents several years before appearing in peer-reviewed journal articles, and in many cases, data in patents may never be published elsewhere. Using a combination of text-mining, chemical structure extraction and filtering methods, a pipeline was developed to identify patents within the SureChEMBL database containing compounds and bioactivity data for understudied targets. Data were manually extracted from the resulting patent documents and included in ChEMBL, illuminating a number of novel targets. This project illustrates the potential value of mining patents whilst highlighting a number of key challenges in creating an efficient and viable workflow.

Beyond journal articles – extracting bioactivity data from patents

256th National Meeting (2018)

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