Screening-based translation of public research encounters painful problems.

Whether identified through high throughput screening or in silico screening, and whether target-based or phenotypic, sets of hits will contain chemical con artists. Such pan-assay interference compounds (PAINS) and other subversive compounds continue to pollute the scientific literature. There are several angles of attack to aid identification of such nonprogressable molecules. One of these rules above all, and this is a demonstration of genuine structure-activity relationships. Recognition of this, which will require a greater effort in medicinal chemistry, will be of general benefit.