QSAR models are widely and successfully used in many research areas. The success of such models highly depends on molecular descriptors typically classified as 1D, 2D, 3D, or 4D. While 3D information is likely important, e. g., for modeling ligand‐protein binding, previous comparisons between the performances of 2D and 3D descriptors were inconclusive. Yet in such comparisons the modeled ligands were not necessarily represented by their bioactive conformations. With this in mind, we mined the PDB for sets of protein‐ligand complexes sharing the same protein for which uniform activity data were reported. The results, totaling 461 structures spread across six series were compiled into a carefully curated, first of its kind dataset in which each ligand is represented by its bioactive conformation. Next, each set was characterized by 2D, 3D and 2D + 3D descriptors and modeled using three machine learning algorithms, namely,k‐Nearest Neighbors, Random Forest and Lasso Regression. Models’ performances were evaluated on external test sets derived from the parent datasets either randomly or in a rational manner. We found that many more significant models were obtained when combining 2D and 3D descriptors. We attribute these improvements to the ability of 2D and 3D descriptors to code for different, yet complementary molecular properties.

Adult polyglucosan body disease (APBD) is a late-onset disease caused by intracellular accumulation of polyglucosan bodies, formed due to glycogen-branching enzyme (GBE) deficiency. To find a treatment for APBD, we screened 1,700 FDA-approved compounds in fibroblasts derived from APBD-modeling GBE1 -knockin mice. Capitalizing on fluorescent periodic acid–Schiff reagent, which interacts with polyglucosans in the cell, this screen discovered that the flavoring agent guaiacol can lower polyglucosans, a result also confirmed in APBD patient fibroblasts. Biochemical assays showed that guaiacol lowers basal and glucose 6-phosphate–stimulated glycogen synthase (GYS) activity. Guaiacol also increased inactivating GYS1 phosphorylation and phosphorylation of the master activator of catabolism, AMP-dependent protein kinase. Guaiacol treatment in the APBD mouse model rescued grip strength and shorter lifespan. These treatments had no adverse effects except making the mice slightly hyperglycemic, possibly due to the reduced liver glycogen levels. In addition, treatment corrected penile prolapse in aged GBE1 -knockin mice. Guaiacol’s curative effects can be explained by its reduction of polyglucosans in peripheral nerve, liver, and heart, despite a short half-life of up to 60 minutes in most tissues. Our results form the basis to use guaiacol as a treatment and prepare for the clinical trials in APBD.

In silico (or virtual) screening has become a common practice in current computer-aided drug design efforts. However, application to hit discovery in the G Protein-Coupled Receptors (GPCRs) arena was until recently hampered by the paucity of crystal structures available for this important class of pharmaceutical targets, forcing practitioners in the field to rely on GPCR models derived either ab initio or through homology modeling approaches. In this work we describe the EPIX in silico screening workflow which consists of the following stages: (1) Target modeling; (2) Preparation of screening library; (3) Docking; (4) Binding mode selection; (5) Scoring; (6) Consensus scoring and (7) Selection of virtual hits. This workflow was applied to the virtual screening of 13 GPCRs (5 biogenic amine receptors, 5 peptide receptors, 1 lipid receptor, 1 purinergic receptor and 1 cannabinoid receptor). Hit rates vary between 4% and 21% with higher hit rates usually obtained for biogenic amines and lower hits rates for peptide receptors. These data are analyzed in the context of the available experimental information (i.e., mutational data), the model (i.e., binding site location, and type of interactions) and the screening library. Specific examples are discussed in more detail as well as the future directions and challenges of this approach to in silico screening.