The three-dimensional structure and function of molecules present many challenges and opportunities for developing an understanding of biological systems. With the increasing availability of molecular structures and the advancing accuracy of structure predictions and molecular simulations, the space for algorithmic advancement on many analytical and predictive problems is both broad and deep. To support this field, a rich set of methods and algorithms are available, addressing a variety of important problems such as protein-protein interactions, the effect of mutations on protein structure and function, and protein structure determination. Despite recent advancements in the field, in particular in protein folding with the development of AlphaFold, many problems still remain unsolved.
In this book we focus on a number of topics in Structural Bioinformatics: Cryo-EM structural detection, protein conformational exploration, elucidation of molecular binding surface using geometry, the effect of mutations, insertions and deletions on protein structural stability, and protein-ligand binding.
Author(s): Nurit Haspel, Filip Jagodzinski, Kevin Molloy
Series: Computational Biology
Publisher: Springer
Year: 2022
Language: English
Pages: 119
City: Cham
Preface
Contents
Protein-Ligand Binding with Applications in Molecular Docking
1 Introduction
1.1 Thermodynamic Basis of Protein-Ligand Interaction
2 Computational Methods for Estimating Binding Free Energy
2.1 Alchemical Methods
2.2 Transition Path Sampling Methods
2.3 End-Point Methods
3 Protein-Ligand Binding Databases
4 Molecular Docking
5 Conclusion
References
Explaining Small Molecule Binding Specificity with Volumetric Representations of Protein Binding Sites
1 Introduction
1.1 Comparison Algorithms for Examining Specificity
2 Specificity Assignment
2.1 Binding Site Representations
2.2 Metrics for Binding Site Comparison
2.3 Comparison Algorithms
2.3.1 Data Structures
2.4 Statistical Models for Binding Site Comparison
3 Component Localization
3.1 Foundations of Structure-Based Component Localization
3.1.1 Comparing Solid Representations with CSG
3.1.2 Solid Representations of Binding Cavities
3.2 Using CSG for Component Localization
3.3 Statistical Models for Component Localization
3.4 Volumetric Alignment
3.5 Flexible Representations for Component Localization
3.5.1 Improving Binding Site Comparison
3.5.2 Flexible Volumetric Comparison of Protein Cavities
3.6 Solid Representations of Electrostatic Isopotentials
4 Discussion
4.1 Future Directions
References
Machine Learning-Based Approaches for Protein Conformational Exploration
1 Introduction
2 Biophysical and Empirical Methods
3 Physics-Based Computational Methods
3.1 Molecular Dynamics
3.2 Monte Carlo Based Search Method
4 Geometric and Robotics-Inspired Methods
4.1 Motion Planning Methods
5 Machine Learning-Based Methods
5.1 Dimensionality Reduction Techniques
5.2 Autoencoders
6 Toolkits for Applying Machine Learning
6.1 Topology and Clustering
6.2 Using a priori Knowledge
7 Conclusions
References
Low Rank Approximation Methods for Identifying Impactful Pairwise Protein Mutations
1 Introduction
2 Related Work
3 Methods
3.1 Phase 1: Generate Exhaustive Pairwise Data
3.2 Phase 2: Sampling Methods
3.3 Phase 3: Smooth Approximation Methods
3.3.1 Singular Value Decomposition Allostery Impact Map
3.3.2 Low Rank Allostery Impact Map
3.3.3 Sparse Plus Low Rank Allostery Impact Map
3.4 Evaluation Metrics
4 Results: SVD Smoothing
4.1 SVD Approximation and Sampling Error
5 Results: Case Study on 2LZM
6 Conclusions and Future Work
References
Detection and Analysis of Amino Acid Insertions and Deletions
1 Introduction
2 Computational Methods of InDel Detection
3 Computational Methods of InDel Analysis
3.1 Machine Learning Based Methods
3.2 Detecting Functional and Fitness Effects of InDels on Protein Structure
3.3 Plasticity of Proteins to InDels
4 Conclusion
References
DeepTracer Web Service for Fast and Accurate De Novo Protein Complex Structure Prediction from Cryo-EM
1 Introduction
2 Procedures
3 Results
3.1 Architecture
3.1.1 Web Application Architecture
3.1.2 Parallel Computing
3.2 Prediction Evaluation
3.3 UI/UX Features
3.3.1 Logger
3.3.2 Visualization
3.3.3 DeepTracer Offline
3.3.4 Email Subscription
3.3.5 Mobile Features
3.3.6 User Engagement
4 Discussion
4.1 Design Philosophy
4.2 Future Features
References
