RNAPosers: Machine Learning Classifiers for Ribonucleic Acid–Ligand Poses

Abstract

Determining the three-dimensional (3D) structures of ribonucleic acid (RNA)−small molecule ligand complexes is critical to understanding molecular recognition in RNA. Computer docking can, in principle, be used to predict the 3D structure of RNA−small molecule complexes. Unfortunately, retrospective analysis has shown that the scoring functions that are typically used for pose prediction tend to misclassify nonnative poses as native and vice versa. Here, we use machine learning to train a set of pose classifiers that estimate the relative "nativeness" of a set of RNA− ligand poses. At the heart of our approach is the use of a pose "fingerprint" (FP) that is a composite of a set of atomic FPs, which individually encode the local "RNA environment" around ligand atoms. We found that by ranking poses based on classification scores from our machine learning classifiers, we were able to recover native-like poses better than when we ranked poses based on their docking scores. With a leave-one-out training and testing approach, we found that one of our classifiers could recover poses that were within 2.5 Å of the native poses in ∼80% of the 80 cases we examined, and, on two separate validation sets, we could recover such poses in ∼60% of the cases. Our set of classifiers, which we refer to as RNAPosers, should find utility as a tool to aid in RNA−ligand pose prediction, and so we make RNAPosers open to the academic community via https://github.com/atfrank/RNAPosers.

Profiling Unbinding Events to Identify Near-Native RNA-Ligand Poses

“…Thus, to explore whether we could bypass simulations and predict ssMD-derived unbinding profiles directly from structure, we trained a structure-based regression model using the partial least squared (PLS) approach. 21 Briefly, for each pose of the five RNA in our dataset, we generated molecular (pose) fingerprints 22 and then paired these features with the exponential fitting parameters from the unbinding profile computed from ssMD simulations that were initialized from that pose. Individual PLS models were then used to predict each parameter ( A, τ , and B ) from the pose features.…”

“…Within the context of such studies, consensus scores can be generated by combining the τ that are rapidly predicted using structure-based regression models with scores from other pose prediction strategies. [22][23][24] One could then use these consensus scores to identify high-confidence poses that can be used as the input for binding energies estimation and other post-processing virtual screening methods.…”

fingeRNAt - a novel tool for high-throughput analysis of nucleic acid-ligand interactions

“…The information on statistics of interactions in RNA-ligand complexes derived from the solved structures can be used to develop bioinformatics methods to predict the structure of such complexes. Methods that enable an analysis of RNA-ligand interactions include docking programs (such as rDock (46,47)) or scoring functions (such as DrugScoreRNA (48), RNAPosers (49), and developed in our laboratory LigandRNA and AnnapuRNA (50,51)). As an output, the aforementioned methods return the proposed binding pose with numerical score(s).…”

RNA–ligand molecular docking: Advances and challenges

“…As another machine‐learning model for RNA‐small molecule binding, RNAPosers 141 contains a set of trained pose classifiers that can estimate the “nativeness” of a ligand for a given structure of the RNA and the ligand. The classifiers are based on the random forest method 242 with an ensemble of 1000 decision trees.…”