Dataset Information

PlayMolecule Glimpse: Understanding Protein-Ligand Property Predictions with Interpretable Neural Networks.

ABSTRACT: Deep learning has been successfully applied to structure-based protein-ligand affinity prediction, yet the black box nature of these models raises some questions. In a previous study, we presented K_DEEP, a convolutional neural network that predicted the binding affinity of a given protein-ligand complex while reaching state-of-the-art performance. However, it was unclear what this model was learning. In this work, we present a new application to visualize the contribution of each input atom to the prediction made by the convolutional neural network, aiding in the interpretability of such predictions. The results suggest that K_DEEP is able to learn meaningful chemistry signals from the data, but it has also exposed the inaccuracies of the current model, serving as a guideline for further optimization of our prediction tools.

SUBMITTER: Varela-Rial A

PROVIDER: S-EPMC8790755 | biostudies-literature | 2022 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

PlayMolecule Glimpse: Understanding Protein-Ligand Property Predictions with Interpretable Neural Networks.

Varela-Rial Alejandro A Maryanow Iain I Majewski Maciej M Doerr Stefan S Schapin Nikolai N Jiménez-Luna José J De Fabritiis Gianni G

Journal of chemical information and modeling 20220103 2

Deep learning has been successfully applied to structure-based protein-ligand affinity prediction, yet the black box nature of these models raises some questions. In a previous study, we presented K<sub>DEEP</sub>, a convolutional neural network that predicted the binding affinity of a given protein-ligand complex while reaching state-of-the-art performance. However, it was unclear what this model was learning. In this work, we present a new application to visualize the contribution of each inpu ...[more]

PMID: 34978201

Similar Datasets

Project description:Protein-ligand interactions (PLIs) are essential for biochemical functionality and their identification is crucial for estimating biophysical properties for rational therapeutic design. Currently, experimental characterization of these properties is the most accurate method, however, this is very time-consuming and labor-intensive. A number of computational methods have been developed in this context but most of the existing PLI prediction heavily depends on 2D protein sequence data. Here, we present a novel parallel graph neural network (GNN) to integrate knowledge representation and reasoning for PLI prediction to perform deep learning guided by expert knowledge and informed by 3D structural data. We develop two distinct GNN architectures: [Formula: see text] is the base implementation that employs distinct featurization to enhance domain-awareness, while [Formula: see text] is a novel implementation that can predict with no prior knowledge of the intermolecular interactions. The comprehensive evaluation demonstrated that GNN can successfully capture the binary interactions between ligand and protein's 3D structure with 0.979 test accuracy for [Formula: see text] and 0.958 for [Formula: see text] for predicting activity of a protein-ligand complex. These models are further adapted for regression tasks to predict experimental binding affinities and [Formula: see text] crucial for compound's potency and efficacy. We achieve a Pearson correlation coefficient of 0.66 and 0.65 on experimental affinity and 0.50 and 0.51 on [Formula: see text] with [Formula: see text] and [Formula: see text], respectively, outperforming similar 2D sequence based models. Our method can serve as an interpretable and explainable artificial intelligence (AI) tool for predicted activity, potency, and biophysical properties of lead candidates. To this end, we show the utility of [Formula: see text] on SARS-Cov-2 protein targets by screening a large compound library and comparing the prediction with the experimentally measured data.

Project description:The extremely large number of unique polymer compositions that can be achieved through copolymerisation makes it an attractive strategy for tuning their optoelectronic properties. However, this same attribute also makes it challenging to explore the resulting property space and understand the range of properties that can be realised. In an effort to enable the rapid exploration of this space in the case of binary copolymers, we train a neural network using a tiered data generation strategy to accurately predict the optical and electronic properties of 350 000 binary copolymers that are, in principle, synthesizable from their dihalogen monomers via Yamamoto, or Suzuki-Miyaura and Stille coupling after one-step functionalisation. By extracting general features of this property space that would otherwise be obscured in smaller datasets, we identify simple models that effectively relate the properties of these copolymers to the homopolymers of their constituent monomers, and challenge common ideas behind copolymer design. We find that binary copolymerisation does not appear to allow access to regions of the optoelectronic property space that are not already sampled by the homopolymers, although it conceptually allows for more fine-grained property control. Using the large volume of data available, we test the hypothesis that copolymerisation of 'donor' and 'acceptor' monomers can result in copolymers with a lower optical gap than their related homopolymers. Overall, despite the prevalence of this concept in the literature, we observe that this phenomenon is relatively rare, and propose conditions that greatly enhance the likelihood of its experimental realisation. Finally, through a 'topographical' analysis of the co-polymer property space, we show how this large volume of data can be used to identify dominant monomers in specific regions of property space that may be amenable to a variety of applications, such as organic photovoltaics, light emitting diodes, and thermoelectrics.

Dataset Information

PlayMolecule Glimpse: Understanding Protein-Ligand Property Predictions with Interpretable Neural Networks.

Publications

PlayMolecule Glimpse: Understanding Protein-Ligand Property Predictions with Interpretable Neural Networks.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets