High Performance Computing for Drug Discovery and Biomedicine (Methods in Molecular Biology, 2716)

✍ Scribed by Alexander Heifetz (editor)

• Publisher: Humana
• Year: 2023
• Tongue: English
• Leaves: 430
• Category: Library

✦ Synopsis

This volume explores the application of high-performance computing (HPC) technologies to computational drug discovery (CDD) and biomedicine. The first section collects CDD approaches that, together with HPC, can revolutionize and automate drug discovery process, such as knowledge graphs, natural language processing (NLP), Bayesian optimization, automated virtual screening platforms, alchemical free energy workflows, fragment-molecular orbitals (FMO), HPC-adapted molecular dynamic simulation (MD-HPC), and the potential of cloud computing for drug discovery. The second section delves into computational algorithms and workflows for biomedicine, featuring an HPC framework to assess drug-induced arrhythmic risk, digital patient applications relevant to the clinic, virtual human simulations, cellular and whole-body blood flow modeling for stroke treatments, prediction of the femoral bone strength from CT data, and many more subjects. Written for the highly successful Methods inMolecular Biology series, chapters include introductions to their respective topics, lists of the necessary software and tools, step-by-step and readily reproducible modeling protocols, and tips on troubleshooting and avoiding known pitfalls.
Authoritative and practical, High Performance Computing for Drug Discovery and Biomedicine allows a diverse audience, including computer scientists, computational and medicinal chemists, biologists, clinicians, pharmacologists and drug designers, to navigate the complex landscape of what is currently possible and to understand the challenges and future directions of HPC-based technologies.

✦ Table of Contents

Dedication
Preface
Contents
Contributors
Chapter 1: Introduction to Computational Biomedicine
1 Introduction
2 Methods and Protocols
2.1 Drug Development
2.2 Personalized Medicine
2.3 Medical Diagnosis with Machine Learning
2.4 Human Digital Twins
3 Summary and Perspective
References
Chapter 2: Introduction to High-Performance Computing
1 Introduction
2 Supercomputer Architectures
3 Supercomputer Components
4 Using a Supercomputer
5 Basics of Parallel Programming
6 Conclusions
References
Chapter 3: Computational Biomedicine (CompBioMed) Centre of Excellence: Selected Key Achievements
1 Introduction
1.1 CompBioMed Overview
1.1.1 Digital Twins
1.2 Computational Biomedicine
1.3 CompBioMed Centre of Excellence
1.3.1 HPC and Supercomputers
1.4 The Healthcare Value Chain
1.4.1 Entrepreneurial Opportunities
1.4.2 Activities
1.4.3 Software
1.4.4 Transforming Industry
1.5 Research
1.5.1 Cardiovascular Medicine
1.5.2 Molecular Medicine
1.5.3 Neuro-musculoskeletal Medicine
1.6 The Clinic
1.6.1 To and from the Clinic
1.6.2 Medical Data
1.7 CompBioMed Partners
1.7.1 Core Partners
1.7.2 Phase 1
1.7.3 Phase 2
1.7.4 Associate Partners
1.8 Biomedical Software: Core Applications
1.8.1 Alya
1.8.2 HemeLB
1.8.3 HemoCell
1.8.4 Palabos
1.8.5 Binding Affinity Calculator
1.8.6 CT2S, ARF10, and BoneStrength
1.8.7 openBF
1.8.8 PlayMolecule
1.8.9 TorchMD
1.8.10 Virtual Assay
2 Selected Key Achievements
2.1 Collaborations
2.1.1 ELEM Biotech
2.2 IMAX Films
2.2.1 The Virtual Humans IMAX Film
2.2.2 The Next Pandemic IMAX Film
2.3 Creating a Culture of HPC Among Biomedical Practitioners
2.4 Free Support to Enable and/or Optimize Applications for HPC
2.5 Providing HPC to Surgery
2.6 FDA-Endorsed Credibility to Biomedical Simulations
References
Chapter 4: In Silico Clinical Trials: Is It Possible?
Abbreviations
1 In Silico Trials Help Solve a Growing Drug Development Challenge
Box 1: Model-Informed Drug Development (MIDD) Landscape
2 A Specific Workflow for In Silico Clinical Trials Powered by Knowledge-Based Modeling
Box 2: Assertion
Box 3: Strength of Evidence (SoE)
3 A Collaborative Knowledge-Based Modeling and In Silico Trial Simulation Software Platform, jinko
3.1 Collaborative White-Box Knowledge Management
3.2 In Silico Clinical Trials Powered by a Distributed Solving Architecture
3.3 Barrierless Analytics and Visualization
3.4 Model Editing and Calibration Tasks
4 Application of a Knowledge-Based In Silico Clinical Trial Approach for the Design of Respiratory Disease Clinical Studies
5 The Future of In Silico Clinical Trials
References
Chapter 5: Bayesian Optimization in Drug Discovery
1 Introduction
2 Bayesian Optimization
2.1 Definition
2.2 BO Process
2.3 Surrogate Model
2.4 Gaussian Process
2.5 Kernel and Input Representation
2.6 Acquisition Function
2.7 Batch and Multiobjective Constraints
2.8 Ranking or Sampling
3 Applications in Drug Discovery
3.1 Hyperparameter Optimization of Machine Learning Models
3.2 Small Molecule Optimization
3.3 Peptide and Protein Sequence Optimization
3.4 Chemical Reaction Condition Optimization
3.5 Small Molecule 3D Conformation Minimization
3.6 Ternary Complex Structure Elucidation
4 Conclusion
References
Untitled
Chapter 6: Automated Virtual Screening
1 Introduction
2 Virtual Screening
2.1 Ligand-Based Virtual Screening
2.2 Shape and Pharmacophore Similarity
2.3 Structure-Based Virtual Screening
2.4 Molecular Docking
2.5 Benchmarking Virtual Screening Methods
2.6 Enrichment
2.7 Receiver Operating Characteristic
2.8 Datasets for Benchmarking
3 The Chemical Space to Explore
3.1 Search Space
3.2 Catalogues of Chemical Suppliers
3.3 Virtual Compounds
3.4 Compound Standardization Pipeline
4 Workflow Systems
4.1 Celery (Python)
4.2 Snakemake
4.3 Apache Airflow
4.4 Microservice Architecture
5 Django and Celery for Automating Virtual Screening
6 Conclusion
References
Chapter 7: The Future of Drug Development with Quantum Computing
Acronyms
1 Introduction
1.1 Computation
1.2 Quantum Computing
1.2.1 Superposition
1.2.2 Entanglement
1.2.3 Qubit/Quantum Bit
1.2.4 Bloch Sphere
1.2.5 Quantum Circuit
1.2.6 Quantum Gates
Pauli Gates (X, Y, Z)
Hadamard Gate (H)
Phase Gates
Controlled Gates
Swap Gate (SWAP)
Toffoli Gate (CCNOT)
1.2.7 Data Encoding Techniques
1.2.8 Results´ Interpretation
1.3 Quantum Annealing
1.4 Hamiltonians
1.5 Physical Implementation
1.6 General Applications
1.7 Limitations of Quantum Compute
1.8 Hybrid Quantum Computing
1.9 Parameterized Circuit
1.10 Variational Quantum Eigensolver
Pseudocode
1.11 The Quantum Approximate Optimization Algorithm (QAOA)
Pseudocode for QAOA:
1.12 Quantum Machine Learning
2 Potential QC Applications to Drug Discovery
2.1 Drug Discovery
2.2 Target Identification
2.2.1 Protein Structure Prediction
2.2.2 Biomarker Identification
2.2.3 Inference of Biological Networks
2.2.4 Single Nucleotide Polymorphism (SNP)
2.2.5 Genome Assembly
2.2.6 Transcription Factor (TF) Binding Analysis
2.3 Target Validation
2.3.1 Protein-Ligand Interaction Simulations
2.3.2 Gene Expression Data Validation
2.3.3 Phylogenetic Tree Inference
2.4 Hit Identification
2.4.1 Quantum-Enhanced Virtual Screening
2.4.2 Molecular Docking Simulations
2.5 Hit-to-Lead Optimization
2.5.1 Quantitative Structure-Activity Relationship (QSAR) Modeling
2.5.2 Pharmocophore Modeling
2.5.3 Drug Design
2.6 Lead Optimization
2.6.1 Multi-target Drug Design
2.6.2 Physicochemical Property Optimization
2.6.3 Lead Design and Optimization
2.7 ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) Prediction
2.7.1 ADMET Modeling
Absorption
Distribution
Metabolism
Excretion
Toxicity
3 Summary
References
Chapter 8: Edge, Fog, and Cloud Against Disease: The Potential of High-Performance Cloud Computing for Pharma Drug Discovery
1 Introduction
1.1 Scientific Application User Persona in Drug Discovery
1.2 Common Types of Scientific App
2 High-Performance Computing Overview
2.1 Best Practices for HPC Scientific Applications
2.2 Container Orchestration
2.3 Type of Infrastructure Based on Scientific App Deployment Paradigms
2.3.1 Standalone Applications
2.3.2 HPC Computing Involves Cluster and Grid Computing
2.3.3 Fat Node High-Performance Computing
2.3.4 Cloud Computing
2.3.5 Fog Computing
2.3.6 Edge Computing
2.4 API Types for Communication in Scientific Apps
2.5 Trends in Cloud Computing-Based Drug Discovery Development
2.6 Ethical Issues in Drug Discovery Using Cloud Computing
3 Summary
References
Chapter 9: Knowledge Graphs and Their Applications in Drug Discovery
1 Introduction
2 Applications of Knowledge Graphs in Drug Discovery
3 Democratizing Access to Biomedical Data
4 Visualizing and Contextualizing Biomedical Data
5 Generating Insights from Knowledge Graphs through Automated Data Mining
6 Machine Learning on Knowledge Graphs for Drug Discovery
7 Transformer Neural Networks for Drug Discovery
8 Explainable AI for Drug Discovery
9 Outlook
References
Chapter 10: Natural Language Processing for Drug Discovery Knowledge Graphs: Promises and Pitfalls
1 Introduction
2 Promises
2.1 Relationship Verbs and/or Causality Between Entities
2.1.1 Example: Protein-Protein Interactions
2.1.2 Case Study: Prioritizing Protein Targets Based on Their Association with a Specific Protein, Cancer, and/or Arthritis
2.2 Other Examples of Expanding KGs Using NLP
2.2.1 Example: Drug-TREATS-Disease
2.2.2 Example: gene-HAS_FEATURE-ProteinFeature
2.2.3 Inclusion of Data Sources Such as Electronic Health Records (EHR)
3 Pitfalls
3.1 Entities Are Incorrectly Identified Leading to Erroneous Relationships
3.2 Relationships Are Wrong Because They Lack Context
3.3 Adding Noise (Assertions Are Not Incorrect But Are Generally Unhelpful Due to Insufficient Granularity)
3.4 Misrepresenting Certainty of Assertion
4 Discussion
5 Conclusion
References
Chapter 11: Alchemical Free Energy Workflows for the Computation of Protein-Ligand Binding Affinities
1 Introduction to AFE
1.1 Recent History
1.2 Using Alchemical Methods to Calculate Relative Binding Free Energies
1.3 Other Binding AFE Methods
2 Introduction to RBFE Workflows
2.1 Running RBFE Simulations
2.2 Components of an RBFE Workflow
2.3 Preparing for an RBFE Workflow-Parameterizing Inputs and Ligand Poses
2.4 Defining the Perturbable Molecule
2.4.1 Topology
2.4.2 Atom Mappings
2.5 Network Generation
2.6 Running the Simulations
2.7 Sampling Methods
2.8 Analysis
3 A Survey of Current RBFE Workflows
3.1 FEW
3.2 FESetup
3.3 FEPrepare
3.4 CHARMM-GUI
3.5 Transformato
3.6 PMX
3.7 QLigFEP
3.8 TIES
3.9 ProFESSA
3.10 PyAutoFEP
3.11 BioSimSpace
3.12 FEP+
3.13 Flare
3.14 Orion
4 The Future for RBFE Workflows
References
Chapter 12: Molecular Dynamics and Other HPC Simulations for Drug Discovery
Abbreviations
1 Introduction
2 HPC and MD
3 Domains of Application
3.1 HPC MD in Drug Discovery
3.1.1 HPC MD Support for the Refinement of Cryo-Electron Microscopy Structures
3.1.2 Special-Purpose HPC MD Simulations-The Anton Machines
3.1.3 HPC MD for Cryptic Pockets
3.1.4 An Alternative to MD: HPC Monte Carlo Simulations
3.1.5 SARS-CoV-2 Studies with HPC MD
3.1.6 The Ultimate Future: Combination of HPC MD and AI/ML
3.2 Protein-Protein Interactions
3.2.1 Conventional Approaches
3.2.2 Artificial Intelligence (AI) Methods
3.2.3 Toward the Simulation of the Cytoplasm
3.3 Virtual Screening
3.3.1 Protein Preparation for Ensemble Docking
3.3.2 Ensemble Docking
3.3.3 Consensus Scoring, Consensus Docking, and Mixed Consensus Scoring
3.3.4 Rescoring and Affinity Calculations
3.3.5 Billion-Compound Databases
3.3.6 Docking Ultra-Large Databases
3.3.7 Deep Docking
4 Conclusion and Outlook
References
Chapter 13: High-Throughput Structure-Based Drug Design (HT-SBDD) Using Drug Docking, Fragment Molecular Orbital Calculations,...
1 Introduction
2 Developing the Input Files
3 Ligand Docking
4 Fragment Molecular Orbitals and FMO-HT
5 Molecular Dynamics
6 The Importance of HPCs
7 The Integration of SBDD Techniques to Develop an Automated Pipeline
References
Chapter 14: HPC Framework for Performing in Silico Trials Using a 3D Virtual Human Cardiac Population as Means to Assess Drug-...
1 Introduction
2 Materials and Methods
2.1 In Vitro Experimentation on Reanimated Swine Hearts
2.1.1 Mechanical and Electrical Data Acquisition
3 Results
3.1 In Silico Experiments
3.2 In Vitro Experiments
4 Discussion
5 Conclusion
References
Chapter 15: Effect of Muscle Forces on Femur During Level Walking Using a Virtual Population of Older Women
1 Introduction
2 Methods
2.1 Participants and Data Acquisition
2.2 Baseline Musculoskeletal Models
2.3 Virtual Population
2.4 Dynamic Simulations and Data Analysis
2.5 Finite Element Model of the Femur
2.6 Static Femoral Loading During Gait and Data Analysis
2.7 Results
2.8 Discussion
References
Chapter 16: Cellular Blood Flow Modeling with HemoCell
1 The Cellular Properties of Blood
2 Methods: Accurate Computational Modeling of Blood Flows
2.1 Simulating Blood on a Cellular Scale
2.2 Simulating Fluid Flow with the Lattice Boltzmann Method
2.3 The Computational Model of the Cells Using the Immersed Boundary Method
2.4 Creating Initial Conditions for Cellular Flow
2.5 Advanced Boundary Conditions
2.6 Performance and Load-Balancing
3 Applications of HemoCell
3.1 Cellular Trafficking and Margination
3.2 Cellular Flow in Microfluidic Devices
3.3 Flow in a Curved Micro-Vessel Section
3.4 Flow of Diabetic Blood in Vessels
References
Chapter 17: A Blood Flow Modeling Framework for Stroke Treatments
1 Introduction
2 Methods
2.1 The Lattice-Boltzmann Method
2.2 Appropriate Boundary Conditions
2.3 Porous Medium Simulation
2.4 Permeability Laws
3 Proof-of-Concept: Minimal Thrombolysis
4 Notes
References
Chapter 18: Efficient and Reliable Data Management for Biomedical Applications
1 Introduction
2 BioMedical Research Data Management
2.1 The FAIR Principles
2.2 Data Formats
2.3 Publication Platforms
2.4 Annotation Schemata
3 Automated Data Management and Staging
3.1 Data Infrastructure in HPC Centers
3.2 File Transfer and Staging Methods in HPC
3.3 EUDAT Components
3.4 The LEXIS Platform and Distributed Data Infrastructure
4 Resilient HPC Workflows with Automated Data Management
4.1 Resilient Workflows
4.2 CompBioMed Workflows on the LEXIS Platform
5 Summary
References
Chapter 19: Accelerating COVID-19 Drug Discovery with High-Performance Computing
1 Introduction
2 Methods and Results
2.1 MD-ML-HPC Workflow
2.1.1 Docking and MD-Based Refinement of Docked Poses
2.1.2 MD-Based Binding Affinity Prediction (with ESMACS or TIES)
2.1.3 ML-Based De Novo Design
3 Results
4 Conclusion
References
Chapter 20: Teaching Medical Students to Use Supercomputers: A Personal Reflection
1 Introduction
2 Course Developments
3 Course Delivery
3.1 Location Location Location
3.2 HPC Resource
4 Challenges and Barriers
5 Future Directions
References
Index