Uncertainty Quantification for Deep Learning Models in Protein-Ligand Binding Affinity Prediction
Explore the source record for details and available documents.
Biological and medical sciences↗
Engineering topics
Jones, Derek
Publications and source records attributed to Jones, Derek.
HDBind: encoding of molecular structure with hyperdimensional binary representations
Traditional methods for identifying “hit” molecules from a large collection of potential drug-like candidates rely on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have a significant limitation in that they require exceptional computing capabilities for even relatively small collections of molecules. Increasingly large and complex state-of-the-art deep learning approaches have gained popularity with the promise to improve the productivity of drug design, notorious for its numerous failures. However, as deep learning models increase in their size and complexity, their acceleration at the hardware level becomes more challenging. Hyperdimensional Computing (HDC) has recently gained attention in the computer hardware community due to its algorithmic simplicity relative to deep learning approaches. The HDC learning paradigm, which represents data with high-dimension binary vectors, allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas (computer vision, bioinformatics, mass spectrometery, remote sensing, edge devices, etc.). To the best of our knowledge, our work is the first to consider HDC for the task of fast and efficient screening of modern drug-like compound libraries. We also propose the first HDC graph-based encoding methods for molecular data, demonstrating consistent and substantial improvement over previous work. We compare our approaches to alternative approaches on the well-studied MoleculeNet dataset and the recently proposed LIT-PCBA dataset derived from high quality PubChem assays. We demonstrate our methods on multiple target hardware platforms, including Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs), showing at least an order of magnitude improvement in energy efficiency versus even our smallest neural network baseline model with a single hidden layer. Our work thus motivates further investigation into molecular representation learning to develop ultra-efficient pre-screening tools. We make our code publicly available at https://github.com/LLNL/hdbind.
59 BASIC BIOLOGICAL SCIENCES↗
Advances in Computational Approaches for Estimating Passive Permeability in Drug Discovery
Passive permeation of cellular membranes is a key feature of many therapeutics. The relevance of passive permeability spans all biological systems as they all employ biomembranes for compartmentalization. A variety of computational techniques are currently utilized and under active development to facilitate the characterization of passive permeability. These methods include lipophilicity relations, molecular dynamics simulations, and machine learning, which vary in accuracy, complexity, and computational cost. This review briefly introduces the underlying theories, such as the prominent inhomogeneous solubility diffusion model, and covers a number of recent applications. Various machine-learning applications, which have demonstrated good potential for high-volume, data-driven permeability predictions, are also discussed. Due to the confluence of novel computational methods and next-generation exascale computers, we anticipate an exciting future for computationally driven permeability predictions.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Clustering Protein Binding Pockets and Identifying Potential Drug Interactions: A Novel Ligand-Based Featurization Method
Not Available
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Accelerators for Classical Molecular Dynamics Simulations of Biomolecules
Atomistic Molecular Dynamics (MD) simulations provide researchers the ability to model biomolecular structures such as proteins and their interactions with drug-like small molecules with greater spatiotemporal resolution than is otherwise possible using experimental methods. MD simulations are notoriously expensive computational endeavors that have traditionally required massive investment in specialized hardware to access biologically relevant spatiotemporal scales. Our goal is to summarize the fundamental algorithms that are employed in the literature to then highlight the challenges that have affected accelerator implementations in practice. We consider three broad categories of accelerators: Graphics Processing Units (GPUs), Field-Programmable Gate Arrays (FPGAs), and Application Specific Integrated Circuits (ASICs). These categories are comparatively studied to facilitate discussion of their relative trade-offs and to gain context for the current state of the art. We conclude by providing insights into the potential of emerging hardware platforms and algorithms for MD.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Discovery of Small-Molecule Inhibitors of SARS-CoV-2 Proteins Using a Computational and Experimental Pipeline
A rapid response is necessary to contain emergent biological outbreaks before they can become pandemics. The novel coronavirus (SARS-CoV-2) that causes COVID-19 was first reported in December of 2019 in Wuhan, China and reached most corners of the globe in less than two months. In just over a year since the initial infections, COVID-19 infected almost 100 million people worldwide. Although similar to SARS-CoV and MERS-CoV, SARS-CoV-2 has resisted treatments that are effective against other coronaviruses. Crystal structures of two SARS-CoV-2 proteins, spike protein and main protease, have been reported and can serve as targets for studies in neutralizing this threat. We have employed molecular docking, molecular dynamics simulations, and machine learning to identify from a library of 26 million molecules possible candidate compounds that may attenuate or neutralize the effects of this virus. The viability of selected candidate compounds against SARS-CoV-2 was determined experimentally by biolayer interferometry and FRET-based activity protein assays along with virus-based assays. In the pseudovirus assay, imatinib and lapatinib had IC 50 values below 10 μM, while candesartan cilexetil had an IC 50 value of approximately 67 µM against M pro in a FRET-based activity assay. Comparatively, candesartan cilexetil had the highest selectivity index of all compounds tested as its half-maximal cytotoxicity concentration 50 (CC 50 ) value was the only one greater than the limit of the assay (>100 μM).
59 BASIC BIOLOGICAL SCIENCES↗
Enabling rapid COVID-19 small molecule drug design through scalable deep learning of generative models
We improved the quality and reduced the time to produce machine learned models for use in small molecule antiviral design. Our globally asynchronous multi-level parallel training approach strong scales to all of Sierra with up to 97.7% efficiency. We trained a novel, character-based Wasserstein autoencoder that produces a higher quality model trained on 1.613 billion compounds in 23 minutes while the previous state of the art takes a day on 1 million compounds. Reducing training time from a day to minutes shifts the model creation bottleneck from computer job turnaround time to human innovation time. Our implementation achieves 318 PFLOPs for 17.1% of half-precision peak. We will incorporate this model into our molecular design loop enabling the generation of more diverse compounds; searching for novel, candidate antiviral drugs improves and reduces the time to synthesize compounds to be tested in the lab.
Jacobs, Sam Ade↗
Improved Protein–Ligand Binding Affinity Prediction with Structure-Based Deep Fusion Inference
Not Available
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Predicting Small Molecule Transfer Free Energies by Combining Molecular Dynamics Simulations and Deep Learning
Accurately predicting small molecule partitioning and hydrophobicity is critical in the drug discovery process. There are many heterogeneous chemical environments within a cell and entire human body. For example, drugs must be able to cross the hydrophobic cellular membrane to reach their intracellular targets, and hydrophobicity is an important driving force for drug–protein binding. Atomistic molecular dynamics (MD) simulations are routinely used to calculate free energies of small molecules binding to proteins, crossing lipid membranes, and solvation but are computationally expensive. Machine learning (ML) and empirical methods are also used throughout drug discovery but rely on experimental data, limiting the domain of applicability. We present atomistic MD simulations calculating 15,000 small molecule free energies of transfer from water to cyclohexane. This large data set is used to train ML models that predict the free energies of transfer. We show that a spatial graph neural network model achieves the highest accuracy, followed closely by a 3D-convolutional neural network, and shallow learning based on the chemical fingerprint is significantly less accurate. A mean absolute error of ~4 kJ/mol compared to the MD calculations was achieved for our best ML model. We also show that including data from the MD simulation improves the predictions, tests the transferability of each model to a diverse set of molecules, and show multitask learning improves the predictions. This work provides insight into the hydrophobicity of small molecules and ML cheminformatics modeling, and our data set will be useful for designing and testing future ML cheminformatics methods.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Binding Affinity Prediction by Pairwise Function Based on Neural Network
In this paper, we present a new approach to estimate the binding affinity from given three-dimensional poses of protein–ligand complexes. In this scheme, every protein–ligand atom pair makes an additive free-energy contribution. The sum of these pairwise contributions then gives the total binding free energy or the logarithm of the dissociation constant. The pairwise contribution is calculated by a function implemented via a neural network that takes the properties of the two atoms and their distance as input. The pairwise function is trained using a portion of the PDBbind 2018 data set. The model achieves good accuracy for affinity predictions when evaluated with PDBbind 2018 and with the CASF-2016 benchmark, comparing favorably to many scoring functions such as that of AutoDock Vina. The framework here may be extended to incorporate other factors to further improve its accuracy and power.
59 BASIC BIOLOGICAL SCIENCES↗