Dr. Hal S. Alper is the Cockrell Family Regents Chair in Engineering #1 at The University of Texas at Austin in the McKetta Department of Chemical Engineering. His research focuses on applying and extending the approaches of metabolic engineering, synthetic biology, systems biology, and protein engineering.

Preface xv 1 From Genome to Actionable Insights in Biotechnology 1James Morrissey, Benjamin Strain, and Cleo Kontoravdi 1.1 Introduction 1 1.2 From Genome to Network 2 1.2.1 Metabolic Networks 3 1.2.1.1 Bottom-Up Approaches for Network Reconstruction 3 1.2.1.2 Top-Down Approaches for Network Reconstruction 4 1.2.2 Networks Beyond Metabolism 5 1.3 From Draft to Functional Network 6 1.3.1 Additional Reactions 6 1.3.1.1 Exchange Reactions 6 1.3.1.2 Demand Reactions 6 1.3.1.3 Transport Reactions 6 1.3.1.4 Spontaneous Reactions 7 1.3.1.5 Nongrowth Associated ATP Maintenance 7 1.3.1.6 Biomass Reaction 7 1.3.2 Network Validation 7 1.3.2.1 Manual Screening 8 1.3.2.2 Screening for Dead-End Reactions and Blocked Metabolites 8 1.3.2.3 Infinite Loops 9 1.3.2.4 Leaks and Siphons 10 1.4 From Functional Network to Model 10 1.4.1 Flux Balance Analysis 11 1.4.2 Flux Variability Analysis 12 1.4.3 Flux Sampling 13 1.5 From Model to In Silico Predictions 15 1.5.1 Constraints 15 1.5.2 Objective Function 16 1.5.3 Validating In Silico Predictions 16 1.5.3.1 Growth Rate Predictions 22 1.5.3.2 Amino Acid Auxotrophies 22 1.5.3.3 Gene Essentialities 22 1.5.3.4 Known Host Traits 23 1.5.3.5 Intracellular Predictive Accuracy 23 1.5.4 Toward Multilayer, Multiscale Metabolic Networks 23 1.5.4.1 Integrating Gene Regulatory Networks 24 1.5.4.2 Integrating Transcription and Translation 25 1.5.4.3 Integrating Signaling Networks 25 1.5.4.4 Multicellular and Multitissue Models 25 1.5.4.5 Multiscale Bioreactor Models 26 1.6 From Predictions to Actionable Insights in Biotechnology 26 1.6.1 Metabolic Engineering 26 1.6.2 Cell Line Development and Metabolic Profiling 27 1.6.3 Media and Feed Design 29 1.6.4 Gene Essentiality 30 1.6.5 Kinetic Parameter Estimation 30 1.6.6 Process Monitoring and Forecasting 30 References 31 2 Automated Approaches for the Development of Genome-Scale Metabolic Network Models 43Emma M. Glass, Deborah A. Powers, and Jason A. Papin 2.1 Introduction 43 2.2 Manual GSM Creation 44 2.3 Automated GSM Development 45 2.3.1 General Approach for Automated GSM Methods 45 2.3.2 GSM Construction Tools 46 2.3.2.1 From Raw Sequences 46 2.3.2.2 From Pre-annotated Sequences 52 2.3.2.3 From Reaction Database Information 53 2.3.2.4 Based on Existing GSMs 56 2.3.2.5 GSM Modification and Visualization Tools 57 2.4 Applications of Automatically-Generated GSM Collections 59 2.4.1 AGORA1 - 773 GSMs 59 2.4.2 EMBL GEMs - 5,587 GSMs 60 2.4.3 MetaGEM - 447 GSMs 61 2.4.4 AGORA2-7,302 GSMs 61 2.4.5 PATHGENN - 914 GSMs 62 2.5 Future Directions for the Field of Automated GSM Development 62 2.6 Conclusion 63 References 63 3 Machine-Guided Approaches for Synthetic Biology Part Design 67Marc Amil, Leandro N. Ventimiglia, and Aleksej Zelezniak 3.1 Introduction 67 3.2 Model-Guided Sequence Design Using Deep Learning 70 3.2.1 Predictive Models for DNA Function: CNNs in Regulatory Sequence Analysis 70 3.2.1.1 Data Considerations for Supervised Learning on Genomic Sequences 72 3.2.1.2 Primer on Convolutional Neural Networks for Supervised Genomic Sequence Modeling 73 3.2.2 Generative Sequence Modeling 75 3.2.2.1 Data Preparation for Unsupervised Learning 76 3.2.2.2 GANs for the Design of Biological Sequences 77 3.2.2.3 Transformer-Based DNA Models 80 3.2.2.4 Diffusion Models for the Design of Biological Sequences 84 3.3 Sources of Sequence-Function Data for Deep Learning 85 3.3.1 Native-Context Genome-Derived Datasets 86 3.3.2 Synthetic Datasets 87 3.4 Evaluating Synthetic Biological Parts Using Motif Analysis and Deep Learning 90 3.5 Current Challenges of Generative Part Design 93 3.6 Conclusion 93 References 94 4 Machine Learning for Sequence-to-Function Approaches 103Rana A. Barghout, Maxim Kirby, Austin Zheng, Lya Chinas, Marjan Mohammadi, Zhiqing Xu, Benjamin Sanchez-Lengeling, and Radhakrishnan Mahadevan 4.1 Introduction 103 4.2 Current State of Sequence-to-Function Modeling 105 4.2.1 Protein Function Prediction: From BLAST to Language Models 105 4.2.2 Gene Ontology 106 4.2.3 Enzyme Commission Numbers 106 4.2.4 Enzyme Activity 109 4.2.5 Protein Thermal Stability 109 4.2.6 Protein Toxicity 110 4.2.7 Protein Solubility 111 4.3 Tool Kits and Benchmarks 111 4.3.1 Overview of Open-Source Tools 111 4.3.2 Importance of Standardized Benchmarks 114 4.4 Emerging ML Methods 115 4.4.1 Contrastive Learning 115 4.4.2 Meta Learning 116 4.5 Case Studies 116 4.5.1 Prediction of Enzyme Activity and Substrate Specificity 116 4.5.1.1 Predicting k cat Using CPI-Pred 117 4.6 Challenges in Sequence-to-Function Mapping 118 4.6.1 Sparse Experimental Data 119 4.6.2 Interpretability 120 4.7 Conclusion and Future Directions 120 References 121 5 Prediction of Enzyme Functions by Artificial Intelligence 131Ha Rim Kim, Hongkeun Ji, Gi Bae Kim, and Sang Yup Lee 5.1 Introduction 131 5.2 Conventional Computational Approaches for Predicting Enzyme Function 132 5.3 Prediction of Enzyme Functions Using Machine Learning 133 5.3.1 Extraction of Enzyme Features from Amino Acid Sequences 134 5.3.2 Machine Learning-Based Approaches Algorithms for Enzyme Function Prediction 136 5.4 Prediction of Enzyme Functions Using Deep Learning 138 5.4.1 Convolutional Neural Network 139 5.4.2 Recurrent Neural Network 141 5.4.3 Transformer and Protein Language Models 142 5.4.4 Graph Neural Network 145 5.5 Concluding Remarks 147 Acknowledgments 153 References 153 6 Design of Biochemical Pathways via AI/ML-Enabled Retrobiosynthesis 161Hongxiang Li, Xuan Liu, and Huimin Zhao 6.1 Introduction 161 6.1.1 Computer-Aided Synthesis Planning 161 6.1.2 Retrobiosynthesis 162 6.2 Retrobiosynthesis Tools 162 6.2.1 Template-Based Tools 162 6.2.2 Template-Free Tools 166 6.2.3 Searching Algorithm 167 6.2.4 Ranking 168 6.3 Enzyme Selection and Optimization 168 6.3.1 Enzyme Substrate Specificity 169 6.3.2 Enzyme Catalytic Efficiency 170 6.3.3 Enzyme Engineering 172 6.3.4 De Novo Enzyme Design and Discovery 172 6.4 Perspectives 174 6.4.1 Integrating Biocatalysis with Chemocatalysis 175 6.4.2 Toward Next-Generation AI for Retrobiosynthesis Planning 175 6.4.3 Enhancing Enzyme Prediction and Design Capabilities 176 6.4.4 Data Standardization and Model Interpretability 177 Acknowledgments 178 References 179 7 Machine Learning to Accelerate the Discovery of Therapeutic Peptides 183Nicole Soto-Garcia, Mehdi D. Davari, and David Medina-Ortiz 7.1 Introduction 183 7.2 Peptides: Definitions and Main Characteristics 184 7.3 Benefits and Limitations of Therapeutic Peptides 185 7.4 Computational Design of Therapeutic Peptides 186 7.5 Data Sources for Peptide Discovery 187 7.6 ML-Based Strategies to Accelerate the Discovery of Therapeutic Peptides 189 7.6.1 Data-Driven Approaches 190 7.6.2 ml Strategies for Peptide Bioactivity Classification 192 7.6.2.1 Classification Models for Antimicrobial Peptides 192 7.6.2.2 Classification Models for Antiviral Peptides 193 7.6.2.3 Classification Models for Antifungal Peptides 194 7.6.2.4 Additional Classification Models for Therapeutic Peptides 194 7.6.3 Strategies for Building Toxicity Classification Models 195 7.6.3.1 Toxicity Prediction 196 7.6.3.2 Immunogenicity Identification 197 7.6.3.3 Hemolysis Evaluation 197 7.6.3.4 Other Toxic Adverse Effect Predictions 198 7.6.4 Data-Driven Strategies for Modeling Pharmacological Profiles 198 7.6.5 De novo Design of Therapeutic Peptides 199 7.6.5.1 Variational Autoencoder-Based Approaches 200 7.6.5.2 Generative Adversarial Networks 200 7.6.5.3 Transformer-Based Language Models 201 7.6.5.4 Diffusion Models 201 7.7 Next-Generation Peptide Design Through Multi-Agent Systems 201 7.8 Developing AI-Agent for Autonomous Therapeutic Peptide Design 203 7.9 Conclusion and Perspectives 205 Acknowledgments 206 References 207 8 Machine Learning Approaches for High-Throughput Microbial Identification/Culturing 219Mohamed Mastouri and Yang Zhang 8.1 Introduction 219 8.2 High-Throughput (HTP) Techniques in Microbial Research 221 8.2.1 Definition and Scope of HTP Microbial Techniques 221 8.2.2 Metagenomics and Next-Generation Sequencing (NGS) 223 8.2.3 Mass Spectrometry-Based Proteomics (MALDI-TOF MS) 223 8.2.4 Flow Cytometry 224 8.2.5 Microfluidics and Lab-on-a-Chip Systems 224 8.2.6 High-Content Imaging and Phenotyping 225 8.3 Fundamentals of Machine Learning 226 8.3.1 Definition of Machine Learning and AI 226 8.3.2 Supervised vs. Unsupervised vs. Reinforcement Learning 226 8.3.3 ML Algorithms Commonly Used in Microbial Identification 229 8.4 Machine Learning Approaches for High-Throughput Microbial Identification 230 8.4.1 Genomic and Metagenomic Data Processing 230 8.4.2 Mass Spectrometry-Based Identification 234 8.4.3 Imaging-Based Identification 236 8.5 Machine Learning Approaches for High-Throughput Microbial Culturing 237 8.5.1 ML-Driven Microbial Growth Prediction 237 8.5.2 AI in Microbial Cultivation Process Optimization 238 8.5.3 Synthetic Biology and AI-Driven Strain Engineering 240 8.6 Challenges and Limitations of Machine Learning in HTP Microbial Research 241 8.7 Future Perspectives and Emerging Trends 242 8.8 Conclusion 243 Acknowledgments 244 References 244 9 Generative AI for Knowledge Mining of Synthetic Biology and Bioprocess Engineering Literature 253Zhengyang Xiao and Yinjie J. Tang 9.1 Introduction 253 9.2 Text Mining Using Knowledge Graph Tools 254 9.2.1 NEKO: A Lightweight Knowledge Graph Tool 254 9.2.2 GraphRAG 256 9.3 LLM-Automated Data Extraction for Machine Learning 258 9.4 Current Limitations 259 9.5 Conclusion 260 Acknowledgments 260 References 260 10 Metabolomics: Big Data Approaches 263Kenya Tanaka, Christopher J. Vavricka, and Tomohisa Hasunuma 10.1 Introduction 263 10.2 Methods for Metabolomics 264 10.2.1 Preparation of Samples 264 10.2.2 Detection and Quantification 266 10.3 Analysis and Application of Metabolomics Data for Biotechnology 267 10.3.1 Metabolomics for Identification of Pathway Bottlenecks 267 10.3.2 Absolute Metabolomics and Thermodynamic Analyses 272 10.3.3 Evaluation and Optimization of Metabolic Flux 272 10.4 Artificial...