Fig. 2From: Machine learning approaches to optimize small-molecule inhibitors for RNA targetingScheme of the study. The study is divided into computational and experimental parts. In the computational part, deep convolutional networks, linear regressor, and classification trees were trained against a dataset of labeled small molecules. In the experimental part, small molecules with features important for RNA target binding were synthesized and then examined for their ability to inhibit ribosome activity (biochemical validation)Back to article page