BitterMatch: recommendation systems for matching molecules with bitter taste receptors

Abstract

Bitterness is an aversive cue elicited by thousands of chemically diverse compounds. Bitter taste may prevent consumption of foods and jeopardize drug compliance. The G protein-coupled receptors for bitter taste, TAS2Rs, have species-dependent number of subtypes and varying expression levels in extraoral tissues. Molecular recognition by TAS2R subtypes is physiologically important, and presents a challenging case study for ligand-receptor matchmaking. Inspired by hybrid recommendation systems, we developed a new set of similarity features, and created the BitterMatch algorithm that predicts associations of ligands to receptors with ~ 80% precision at ~ 50% recall. Associations for several compounds were tested in-vitro, resulting in 80% precision and 42% recall. The encouraging performance was achieved by including receptor properties and integrating experimentally determined ligand-receptor associations with chemical ligand-to-ligand similarities.BitterMatch can predict off-targets for bitter drugs, identify novel ligands and guide flavor design. The novel features capture information regarding the molecules and their receptors, which could inform various chemoinformatic tasks. Inclusion of neighbor-informed similarities improves as experimental data mounts, and provides a generalizable framework for molecule-biotarget matching. Graphical Abstract

Inhibiting a promiscuous GPCR: iterative discovery of bitter taste receptor ligands

“…The only two molecules with at least one predicted protein target - no TAS2R - with a probability higher than 0.25 are LW118 (max value=0.42) and LW209 (max value=0.22), while the average probability for all of the 292 potential targets found by the algorithm for the 14 antagonists is limited to 0.11. In addition, all the ligands found within this study underwent BitterMatch, a machine learning algorithm for prediction of TAS2R subtypes binding bitter molecules with 80% precision in prospective testing [56]. The majority (131 out of 210, ~62%) of the compounds are predicted to be selective for TAS2R14.…”

“…BitterMatch algorithm was applied to the list of agonists and antagonists according to the protocol described in [54]. Briefly, the 3D structures of the compounds were prepared using LigPrep (Schrödinger Release 2021-1: LigPrep, Schrödinger, LLC, New York, NY, 2021) and Epik (Schrödinger Release 2021-1: Epik, Schrödinger, LLC, New York, NY, 2021) in pH = 7.0±0.5.…”

“…BitterMatch algorithm was applied to the list of agonists and antagonists according to the protocol described in [54]. Briefly, the 3D structures of the compounds were prepared using LigPrep (Schrödinger…”

“…2B) and, therefore, affecting the activation status of the receptor. The selectivity of the new ligands was evaluated using BitterMatch, a machine learning algorithm for prediction of TAS2R subtypes associated with bitter molecules [54]. BitterMatch was shown to have 80% precision in prospective testing.…”

Potential role of the bitter taste receptor T2R14 in the prolonged survival and enhanced chemoresponsiveness induced by apigenin

“…In particular, apigenin has been reported to activate another T2R subtype, T2R39 ( 28 ), which is not expressed in PDAC tissue according to our own not yet published data. Recently, the effects of apigenin were assessed on all human T2Rs; this previous study confirmed that it could activate T2R14, but not T2R39, and identified T2R43 as the only additional T2R activated by apigenin ( 50 ). However, T2R43 was activated by 30 µ M apigenin and not by 10 µ M, which is the concentration that was used in the present study.…”