Deep Learning for the Life Sciences

Book description

Deep learning has already achieved remarkable results in many fields. Now it’s making waves throughout the sciences broadly and the life sciences in particular. This practical book teaches developers and scientists how to use deep learning for genomics, chemistry, biophysics, microscopy, medical analysis, and other fields.

Ideal for practicing developers and scientists ready to apply their skills to scientific applications such as biology, genetics, and drug discovery, this book introduces several deep network primitives. You’ll follow a case study on the problem of designing new therapeutics that ties together physics, chemistry, biology, and medicine—an example that represents one of science’s greatest challenges.

  • Learn the basics of performing machine learning on molecular data
  • Understand why deep learning is a powerful tool for genetics and genomics
  • Apply deep learning to understand biophysical systems
  • Get a brief introduction to machine learning with DeepChem
  • Use deep learning to analyze microscopic images
  • Analyze medical scans using deep learning techniques
  • Learn about variational autoencoders and generative adversarial networks
  • Interpret what your model is doing and how it’s working

Publisher resources

View/Submit Errata

Table of contents

  1. Preface
    1. Conventions Used in This Book
    2. Using Code Examples
    3. O’Reilly Online Learning
    4. How to Contact Us
    5. Acknowledgments
  2. 1. Why Life Science?
    1. Why Deep Learning?
    2. Contemporary Life Science Is About Data
    3. What Will You Learn?
  3. 2. Introduction to Deep Learning
    1. Linear Models
    2. Multilayer Perceptrons
    3. Training Models
    4. Validation
    5. Regularization
    6. Hyperparameter Optimization
    7. Other Types of Models
      1. Convolutional Neural Networks
      2. Recurrent Neural Networks
    8. Further Reading
  4. 3. Machine Learning with DeepChem
    1. DeepChem Datasets
    2. Training a Model to Predict Toxicity of Molecules
    3. Case Study: Training an MNIST Model
      1. The MNIST Digit Recognition Dataset
      2. A Convolutional Architecture for MNIST
    4. Conclusion
  5. 4. Machine Learning for Molecules
    1. What Is a Molecule?
      1. What Are Molecular Bonds?
      2. Molecular Graphs
      3. Molecular Conformations
      4. Chirality of Molecules
    2. Featurizing a Molecule
      1. SMILES Strings and RDKit
      2. Extended-Connectivity Fingerprints
      3. Molecular Descriptors
    3. Graph Convolutions
    4. Training a Model to Predict Solubility
    5. MoleculeNet
      1. SMARTS Strings
    6. Conclusion
  6. 5. Biophysical Machine Learning
    1. Protein Structures
      1. Protein Sequences
      2. A Short Primer on Protein Binding
    2. Biophysical Featurizations
      1. Grid Featurization
      2. Atomic Featurization
    3. The PDBBind Case Study
      1. PDBBind Dataset
      2. Featurizing the PDBBind Dataset
    4. Conclusion
  7. 6. Deep Learning for Genomics
    1. DNA, RNA, and Proteins
    2. And Now for the Real World
    3. Transcription Factor Binding
      1. A Convolutional Model for TF Binding
    4. Chromatin Accessibility
    5. RNA Interference
    6. Conclusion
  8. 7. Machine Learning for Microscopy
    1. A Brief Introduction to Microscopy
      1. Modern Optical Microscopy
    2. The Diffraction Limit
      1. Electron and Atomic Force Microscopy
      2. Super-Resolution Microscopy
      3. Deep Learning and the Diffraction Limit?
    3. Preparing Biological Samples for Microscopy
      1. Staining
      2. Sample Fixation
      3. Sectioning Samples
      4. Fluorescence Microscopy
      5. Sample Preparation Artifacts
    4. Deep Learning Applications
      1. Cell Counting
      2. Cell Segmentation
      3. Computational Assays
    5. Conclusion
  9. 8. Deep Learning for Medicine
    1. Computer-Aided Diagnostics
    2. Probabilistic Diagnoses with Bayesian Networks
    3. Electronic Health Record Data
      1. The Dangers of Large Patient EHR Databases?
    4. Deep Radiology
      1. X-Ray Scans and CT Scans
      2. Histology
      3. MRI Scans
    5. Learning Models as Therapeutics
    6. Diabetic Retinopathy
    7. Conclusion
      1. Ethical Considerations
      2. Job Losses
      3. Summary
  10. 9. Generative Models
    1. Variational Autoencoders
    2. Generative Adversarial Networks
    3. Applications of Generative Models in the Life Sciences
      1. Generating New Ideas for Lead Compounds
      2. Protein Design
      3. A Tool for Scientific Discovery
      4. The Future of Generative Modeling
    4. Working with Generative Models
      1. Analyzing the Generative Model’s Output
    5. Conclusion
  11. 10. Interpretation of Deep Models
    1. Explaining Predictions
    2. Optimizing Inputs
    3. Predicting Uncertainty
    4. Interpretability, Explainability, and Real-World Consequences
    5. Conclusion
  12. 11. A Virtual Screening Workflow Example
    1. Preparing a Dataset for Predictive Modeling
    2. Training a Predictive Model
    3. Preparing a Dataset for Model Prediction
    4. Applying a Predictive Model
    5. Conclusion
  13. 12. Prospects and Perspectives
    1. Medical Diagnosis
    2. Personalized Medicine
    3. Pharmaceutical Development
    4. Biology Research
    5. Conclusion
  14. Index

Product information

  • Title: Deep Learning for the Life Sciences
  • Author(s): Bharath Ramsundar, Peter Eastman, Pat Walters, Vijay Pande
  • Release date: April 2019
  • Publisher(s): O'Reilly Media, Inc.
  • ISBN: 9781492039839