Activity profiles of analog series containing pan assay interference compounds

Abstract

Activity artifacts in assays present a major problem for biological screening and medicinal chemistry. Such artifacts are often caused by compounds that form aggregates or are reactive under assay conditions. Many pan assay interference compounds (PAINS) have been proposed to cause false-positive assay readouts. PAINS are typically contained as substructures in larger molecules. They are used as computational filters to detect compounds with potential chemical liabilities. Recent studies have shown that molecules containing the same PAINS substructure often have greatly varying hit rates in screening assays. Even the overall most frequently active PAINS substructures are found in compounds that are only rarely active or consistently inactive in many assays they are tested in. These observations suggest that the structural context in which PAINS are presented may play an important role for eliciting false-positive activities. However, this assumption remains to be investigated. Herein, we report the systematic identification of analog series of screening compounds that contain PAINS or exclusively consist of PAINS and the analysis of their activity profiles. Comparison of analogs or different series of analogs containing the same PAINS substructure provides structural context information. For many PAINS, extensively tested series with distinct activity profiles were detected. Furthermore, analogs within the same series often displayed significant differences in hit rates. The analog series reported herein organize PAINS in different structural contexts. Their activity profiles provide many opportunities for experimental follow-up investigations to better understand PAINS characteristics.