Nature Breakthrough: AlphaFold 3 Deciphers the 'Book of Life' After Two Decades, Solving Biology’s Greatest Challenge

Jul 01, 2024 08:12 CST Updated 08:12

Isomorphic Labs

AI Drug Developer

May 8, 2024, GoogleDeepMindandIsomorphic LabsJointly Released on *Nature*: The Latest AI Model AlphaFold 3 in the Protein Field! This model can accurately predict the structures of proteins, DNA, RNA, and ligands, as well as their interaction patterns.

This is another major breakthrough following AlphaFold 2 in the prediction of drug-like interactions. AlphaFold 3 has achieved unprecedented accuracy, including the binding of proteins with ligands and the binding of antibodies with their target proteins. In the PoseBusters benchmark, AlphaFold 3's accuracy is 50% higher than the best existing traditional methods, and it requires no structural information input, making it the first artificial intelligence system to surpass traditional physics-based prediction tools. The ability to predict antibody-protein binding is crucial for understanding various aspects of human immune responses and designing new antibodies.

In recent years, innovations in biotechnology have led to the accumulation of omics data at an astonishing rate, ushering in the "big data" era. Extracting inherent valuable knowledge from various types of omics data remains a daunting challenge in bioinformatics. Better solutions often require more innovative approaches for efficient processing and effective outcomes. Recent advances in the integrative analysis of multi-omics data and computational modeling have helped address these needs in increasingly harmonious ways. The development and application of machine learning have significantly advanced our understanding of biology and biomedicine, greatly facilitating the development of therapeutic strategies, particularly in precision medicine. Here, we propose a comprehensive investigation and discussion of what has happened, what is happening, and what will happen when machine learning meets omics. Specifically, we describe how artificial intelligence is applied in omics research and review the latest advances at the interface between machine learning and the broadest range of omics, including genomics, transcriptomics, proteomics, metabolomics, radiomics, and single-cell omics.

Deep Learning Protein Design

CADD Computer-Aided Drug Design

AIDD Artificial Intelligence Drug Discovery and Design

Deep Learning Genomics

Machine Learning Microbiomics

Machine Learning Metabolomics

CRISPR-Cas9 Gene Editing Technology

Single-cell Sequencing and Spatial Multi-omics

The following is a content introduction

WORK OVERVIEW

Deep Learning Protein Design

Content can be scrolled down

Day One

1. The Importance of Deep Learning in Protein Design

1.1 Overview of the Basic Concepts of Deep Learning

What is Deep Learning

Basic Concepts of Neural Networks

The History and Development of Deep Learning

1.2 Application of Deep Learning in Bioinformatics

Examples of applications in genomics, bio-component modification, and protein design

2. Basic Concepts of Protein Structure and Function

2.1 From Primary to Higher Levels of Protein StructureQuaternary Structure

Amino acid sequence, secondary structure, tertiary structure, quaternary structure

2.2 Relationship between Structure and Function

The Principle of Structure Determines Function (Anfinsen's Principle)

Specific Examples of Protein Functions

2.3 Introduction to Important Protein Databases

Protein Sequence Database

Protein Structure Database (PDB Database, etc.)

2.4 Introduction to Data Representation of Proteins

Protein Sequence Representation

Protein Structure Representation (Including Practical Demonstrations: Protein Structure Manipulation Using BioPython, Pymol, etc.)

The First Day, Afternoon

1. Introduction to Deep Learning Basics

1.1 Introduction to Neural Networks

Basic Components of Neural Networks

Forward Propagation and Backward Propagation

1.2 Common Deep Learning Algorithms (Convolutional Neural Networks, Recurrent Neural Networks, etc.)

Convolutional Neural Network (CNN)

Recurrent Neural Network (RNN)

Long Short-Term Memory (LSTM)

Transformer Structure

Diffusion Model

2. Deep Learning Tools and Frameworks

2.1 TensorFlow Framework

Introduction to the Basic Concepts of TensorFlow

2.2 PyTorch Framework (Key Focus)

Basic Concepts of PyTorch

Basic Usage of PyTorch (Including Practical Demonstration: Building Neural Networks with PyTorch)

The Next Day

1. Introduction to Protein Sequence Analysis

1.1 Sequence Alignment Tools and Methods

Introduction to Tools such as BLAST and Clustal Omega (Including Practical Demonstrations: How to Use These Tools)

The Importance of Multiple Sequence Alignment (MSA)

Protein Sequence Similarity Analysis Methods Based on Mathematics or Deep Learning (Including Practical Demonstration: Demonstration of Sequence Feature Extraction Methods)

1.2 Application of Deep Learning in Protein Sequence Classification

Automatic Feature Extraction

Protein Family Classification

Protein Function Classification

2. Introduction to Protein Large Language Models (Key Focus)

2.1 Introduction to Language Models

What is Natural Language Processing (NLP)?

What are Large Language Models

2.2 Language Models in Biology

Sequence Language Model kmer2vec, DNABert (Including Practical Demonstration: Simple Training of a kmer2vec Model for Protein Sequences)

Protein Large Language Models (e.g., ESM2)

The Next Afternoon

1. Detailed Introduction to Various Knowledge of Natural Language Processing (NLP)

1.1 Word Embedding

One-hot embedding

Word2vec embedding

Glove embedding

Language Large Model Embedding

1.2 Introduction to Commonly Used Loss Functions in Deep Learning

2. Deep Learning Model Optimization Techniques

2.1 Model Evaluation and Optimization Methods

Evaluation Metrics: Accuracy, Recall, F1 Score, etc.

Optimization methods: Regularization, Dropout, etc.

2.2 Hyperparameter Tuning

The Impact of Hyperparameters on the Model

Grid Search, Random Search

Day 3

1. Basic Concepts of Protein Design

1.1 Overview of Protein Engineering and Design

History of Protein Engineering

Methods of Protein Design

1.2 Challenges and Opportunities in Protein Design

The Challenge of Designing New Functional Proteins

Prospects for the Application of Protein Design

2. Introduction to the Central Dogma of Protein Design

2.1 Positive Central Dogma

2.2 Reverse Central Dogma

The Third Day, Afternoon

1. Application of Deep Learning in Protein Structure Prediction (Key Focus)

1.1 The Development Process of Protein Structure Prediction

Summary of David Baker's Research Work

Conversion Method between Protein 3D Coordinates and Internal Protein Coordinates

1.2 AlphaFold and Its Impact

How AlphaFold 2 Works

OpenFold Code Introduction

Differences Between AlphaFold 3 and AlphaFold 2 (Including Practical Demonstration: Protein Structure Prediction Using AlphaFold 3)

1.3 Other Deep Learning Models

Introduction to Other Models such as RosettaFold

Day 4

1. Application of Deep Learning in Protein Function Prediction for Drug Discovery

1.1 Overview of Deep Learning and Drug Design

The Importance of Deep Learning in Drug Design

1.2 Protein-Small Molecule Interaction Prediction

Introduction to Various Cutting-edge Prediction Methods and Models

2. Protein Function Prediction: Protein-Protein Interactions

2.1 Model Selection and Training

Introduction to the Characteristics of Different Models

2.2 Evaluation and Interpretation of Prediction Results

Evaluation Indicators and Methods

2.1 Introduction to Protein-Protein Interactions After Point Mutations

Examples of Deep Learning Models Playing a Role in Coronavirus Prediction

The Fourth Day, Afternoon

1. Protein Function Prediction: Protein-Nucleic Acid Interactions

1.1 Protein-DNA Interaction Task

1.2 Protein-RNA Interaction Tasks

1.2 Model Selection and Training

Introduction to the Characteristics of Different Models

2. Molecular Docking Combined with Deep Learning

2.1 Basic Concepts of Molecular Docking

The Principle of Molecular Docking

Introduction to Classic Molecular Docking Software Tools (Including Practical Demonstrations: How to Use Various Molecular Docking Software Tools Such as AutoDock, and Presenting Docking Results Using Tools like Pymol)

2.2 Challenges in Molecular Docking

The Challenge of Computational Complexity

2.3 Application of Deep Learning in Molecular Docking

Introduction of New Methods and New Tools

Improving Docking Efficiency with Deep Learning

Day Five

1. Design of Peptide-Protein Interactions

1.1 Basic Concepts of Peptide Design

Characteristics and Design Methods of Polypeptides

1.2 Application of Deep Learning in Peptide Design

Application Case Analysis

2. Reverse Central Dogma: Inferring Structure from Function

2.1 Division of Protein Functional Motifs

2.2 Introduction to the Method of Deriving Required Sequences from Functions

The Fifth Day, Afternoon

1. Reverse Central Dogma: Designing Structures from Scratch

2. Reverse Central Dogma: Design Sequences to Define Structure and Function

3. Summary and Outlook

3.1 The Future Development of Deep Learning in Protein Design

Future Development Trends

The Impact of Protein Foundation Models (Including Practical Demonstrations: How to Apply Foundation Models, How to Design Downstream Tasks Based on Foundation Models)

3.2 Course Summary and Review

Review and Summary of Course Content

4. Q&A Session

CADD Computer-Aided Drug Design

Content can be scrolled down

Day One

Introduction and Fundamentals

1. The Importance of Protein 3D Structure Prediction for Drug Discovery

1.1 Homology Modeling

1.2 De Novo Modeling

2. The Importance of Protein (Enzyme/Target) Active Sites in Drug Discovery

3. Key Structural Features in Drug Discovery (Especially Small Molecules)

4. Common Computational Methods for Drug Discovery

4.1 Molecular Docking

4.2 Virtual Screening

4.3 Molecular Dynamics Simulation

4.4 Others

Introduction to the PDB Database

1.1 Retrieval Protein

1.2 Page Functions and Interpretation

1.3 Data Download

1.4 Interpretation of the PDB File Format

2. PyMol

2.1 Software Introduction

2.2 Introduction to Basic Operations

2.3 Protein and Small Molecule Surface Diagrams, Electrostatic Potential Representation

2.4 Plotting Interaction Diagrams and Creating Simple Animations

The Next Day

Homology Modeling

1. Introduction to Homology Modeling Principles

1.1 Functions and Use Cases of Homology Modeling

1.2 Methods of Homology Modeling

2. Swiss-Model Homology Modeling;

2.1 Search for Homologous Proteins (BLAST, etc.)

2.2 Protein Sequence Alignment

2.3 Protein Template Selection

2.4 Protein Model Construction

2.5 Model Evaluation (Protein Raman Spectra)

2.6 Protein Model Optimization

Case Study and Practice: Modeling with the 2019-nCoV Spike Protein Sequence and Evaluating the Model Based on Corresponding Parameters and Methods

Small Molecule Construction

1. Introduction to ChemDraw Software

1.1 Small Molecule Structure Construction

1.2 Calculation of Small Molecule Physicochemical Properties (e.g., Molecular Weight, clogP, etc.)

1.3 Construct macromolecules, amino acids, DNA, RNA, and other molecules separately

Small Molecule Compound Library

2 Small Molecule Database

2.1 DrugBank、ZINC、ChEMBLIntroduction and Usage of Databases

2.2 Introduction and Use of Natural Product and Traditional Chinese Medicine Component Databases

Day 3

Basics of Molecular Docking

1.1 Principle of Molecular Docking

1.2 Classification of Molecular Docking

1.3 Molecular Docking Scoring Function

2. Practice of Conventional Molecular Docking

2.1 Implementation of Docking

2.1.1 Preparation of Drug Molecular Ligands

2.1.2 Preparation of Protein Receptors

2.1.3 Receptor Grid Calculation

2.1.3 Perform Semi-Flexible Docking

Result Evaluation

1.2.1 Comparison of Crystal Structure Conformations

1.2.2 Evaluation of Docking Results from an Energy Perspective

1.2.3 Evaluation of Docking Results by Cluster Analysis

1.2.4 Selection of the Optimal Binding Conformation

2 Integration with Other Methods

Day 4

1 Flexible Docking

1.1 Small Molecule Ligand Optimization Preparation

1.2 Preparation of Protein Receptors

1.3 Definition of Flexible Residues

1.4 Protein Receptor Grid Calculation

1.5 Flexible Docking Calculation and Result Evaluation

1.6 Comparison and Selection between Semi-flexible Docking and Flexible Docking

2 Flexible docking implementation through other methods

Afternoon

Receptor-Based Drug Discovery

1 Preparation for Virtual Screening

1.1 Different Formats of Small Molecule Files

1.2 Introduction to the Most Practical Features of OpenBabel

1.3 Conversion of Small Molecules in Different Formats

2. Docking-Based Virtual Screening

2.1 Definition, Workflow Construction, and Demonstration of Virtual Screening

2.2 Target Protein Selection, Compound Library Acquisition

2.3 Virtual Screening

2.4 Result Analysis (Scoring, Energy, and Interaction Analysis)

Day 5

Morning

Some special molecular docking

1.Small molecule-small molecule docking

1.1 Introduction to Small Molecule-Small Molecule Interactions

1.2 Small Molecule Structure Preprocessing

1.3 Small Molecule-Small Molecule Docking (Sugar-Small Molecule as an Example)

1.4 Display and Analysis of Docking Results

2. Protein-Nucleic Acid Docking

3. Protein-Protein Docking

Afternoon

Ligand-Based Drug Discovery

1. Construction of 3D-QSAR Model (Sybyl Software)

1.1 Small Molecule Construction

1.2 Creation of Small Molecule Database

1.3 Small Molecule Charge Addition and Energy Optimization

1.4 Determination and Superposition of Molecular Active Conformations

1.5 Creation of 3D-QSAR Model

1.6 Construction of CoMFA and CoMSIA Models

1.7 Test Set Validation Model

1.8 Model Parameter Analysis

1.9 Model Equipotential Map Analysis

1.10 3D-QSAR Model-Guided Drug Design

Day 6

Morning

1. Introduction to Linux System

2. Introduction to Common Commands

3. Installation of Programs on Linux (GROMACS)

Afternoon

MD Practice 1: Molecular Dynamics Simulation of Protein under Solvation

Fully familiar with the general process of molecular dynamics simulations

Day Seven

Morning

MD Practice II: Molecular Dynamics Simulation of Protein-Ligand under Solvation

Mastering Force Field Fitting for Non-Standard Residues

Afternoon

Commonly Used Analysis Commands in Molecular Dynamics Simulations

Calculation of Protein-Ligand Binding Free Energy

Partial Model Case Images

AIDD Artificial Intelligence-Aided Drug Discovery and Design

Content can be scrolled down

Day One

1 Introduction to Artificial Intelligence Drug Discovery (AIDD)

2 Application of Machine Learning and Deep Learning in the Field of Drug Discovery

2.1 Molecular Property Prediction and Optimization

2.2 Virtual Screening

2.3 Drug Side Effect Prediction and Safety Assessment

2.4 New Drug Molecular Design

3 Introduction and Installation of Tools

3.1 Installation of Anaconda3/Pycharm

3.2 Numpy Basics

3.3 Pandas Basics

3.4 Matplotlib Basics

3.5 Scikit-learn Basics

3.6 Pytorch Basics

3.7 RDKit Basics

The Next Day

1 Introduction to Machine Learning

1.1 The Four Elements of Machine Learning

1.2 Data Module

1.3 Core and Advanced APIs

2 Regression Algorithms and Applications

2.1 Linear Regression

2.2 Lasso Regression

2.3 Ridge Regression

2.4 ElasticNset Elastic Network

3 Classification Algorithms and Applications

3.1 Logistic Regression

3.2 Naive Bayes

3.3 KNN

3.4 SVC

3.5 Decision Tree

3.6 Random Forest

3.7 Integrated Learning

4 Clustering Algorithm

4.1 KMeans

4.2 Density-Based Clustering DBSCAN

5 Dimensionality Reduction

5.1 Singular Value Decomposition SVD

5.2 Principal Component Analysis PCA

5.3 Non-negative Matrix Factorization NMF

6 Evaluation Methods and Metrics of the Model

6.1 Hyperparameter Optimization

6.2 Cross-validation

6.3 Evaluation Metrics

7 Feature Engineering

8 Machine Learning Drug Discovery Case Study (I)

——Compound Bioactivity Classification Model

9 Machine Learning Drug Discovery Case Studies (II)

——Compound Bioactivity Regression Model

10 Machine Learning Drug Discovery Cases (III)

—— Drug Side Effect Prediction Model

Figure 1. Side effects propagate in the drug-drug similarity networkBroadcast.

Day 3

1 Deep Learning and Drug Discovery (I)

1.1 Deep Neural Networks

1.2 Forward and Backward Propagation

1.3 Optimization Methods

1.3.1 Gradient Descent with Momentum

1.3.2 Adaptive Learning

1.3.3 Adam

1.4 Loss Function

1.4.1 Mean Absolute Error

1.4.2 Mean Squared Error Loss Function

1.4.3 Cross-Entropy Loss Function

1.5 Convolutional Neural Network

1.5.1 Convolutional Layer

1.5.2 Padding and Stride

1.5.3 Pooling Layer

1.5.4 LeNet Network

1.5.5 AlexNetNetwork

2 Deep Learning Drug Discovery Case Study (I)

—— Drug-Drug Interaction Prediction Model

Figure2. Utilizing drug association network characteristics and deep neural networks to identify potential interaction events between drugs.

Day 4

1 Deep Learning and Drug Discovery (II)

1.1 Recurrent Neural Network

1.2 Message Passing Neural Networks

1.3 Graph Convolutional Neural Networks

1.4 Graph Attention Neural Network

1.5 Graph Sampling and Aggregation

2 Deep Learning Drug Discovery Case Studies (II)

—— Drug Target Interaction Prediction Model

3 Deep Learning Drug Discovery Case Studies (III)

—— Drug Repositioning Model

Day 5

1 Deep Learning and Drug Discovery (III)

1.1 Attention Mechanism

1.2 Self-Attention Model

1.3 Multi-head Self-Attention Model

1.4 Cross-Attention Model

1.5 Transformer Model

2 Deep Learning Drug Discovery Case Studies (IV)

—— Drug-Drug Interaction Prediction Model

3 Deep Learning Drug Discovery Cases (V)

—— Drug Target Binding Affinity Prediction Model

Figure 4.Linking drug and protein information with attention blocksAttentionDTAModel

Deep Learning Genomics

Content can be scrolled down

Day 1

Theoretical Part

Introduction to Deep Learning Algorithms

1. Neural Network Algorithms for Supervised Learning

1.1 Examples of Fully Connected Deep Neural Networks (DNN) in Genomics

1.2 Application Examples of Convolutional Neural Networks (CNN) in Genomics

1.3 Application Examples of Recurrent Neural Networks (RNN) in Genomics

1.4 Application Examples of Graph Convolutional Network (GCN) in Genomics

2. Unsupervised Neural Network Algorithms

2.1 Application Examples of Autoencoder AE in Genomics

2.2 Examples of the Application of Generative Adversarial Networks (GAN) in Genomics

Practical Content

1. Linux Operating System

1.1 Commonly Used Linux Commands

1.2 Vim Editor

1.3 Genomic Data File Management, Modify File Permissions

1.4 Explore Genomic Regions

2. Python Language Basics

2.1.Python Package Installation and Environment Setup

2.2. Common Data Structures and Data Types

The Next Day

Fundamentals of Genomics

1. Genome Database

2. Epigenome

3. Transcriptome

4. Proteome

5. Functional Genomics

Practical Content

Common Deep Learning Frameworks in Genomics

1. Install and introduce deep learning toolkits TensorFlow, Keras, PyTorch

2. Identify deep learning model elements in the toolkit

2.1. Data Representation

2.2. Tensor Operations

2.3. "Layers" in Neural Networks

2.4. Model Composed of Layers

2.5. Loss Function and Optimizer

2.6. Dataset Splitting

2.7. Overfitting and Underfitting

3. Genomic Data Processing

3.1 Install and use keras_dna to process various genomic sequence data such as BED,GFF, GTF, BIGWIG, BEDGRAPH, WIG, etc.

3.2 Designing Deep Learning Models Using keras_dna

3.3 Using keras_dna to Split Training Set and Test Set

3.4 Using keras_dna to select gene sequences of specific chromosomes, etc.

Day 3

Commonly Used Analysis Methods for Microbiome (Practical Operations)

1. Microbial Abundance Analysis

2. Transcriptome Abundance Analysis

3. Phylogenetic Tree Analysis

4. Dimensionality Reduction Analysis

Day 4

Theoretical Part

Deep Learning in Identifying Copy Number Variations DeepCNV

1. Prediction of Copy Number Variations (CNV) in SNP Microarrays, DeepCNV

2. Prediction of premiRNA in RNA-Seq, dnnMiRPre

Practical Content

1. Reproduce DeepCNV Utilizing SNP Microarray Combined with Image Analysis to Identify Copy Number Variations

Reproducing Recurrent Neural Network RNN Tool dnnMiRPre for Predicting pre-miRNA from RNA-Seq

Day Five

Theoretical Part

Application of Deep Learning in Identification and Disease Phenotypes and Biomarkers

1. DeepType: A Deep Learning Tool for Identifying Breast Cancer Subtypes from Gene Expression Data

Practical Content

1. Reproduce DeepType to distinguish breast cancer subtypes from METABRIC breast cancer data.

Day Six

Theoretical Part

Application of Deep Learning in Predicting Drug Response Mechanisms

1. SWnet: A Deep Learning Tool for Predicting Drug Response Mechanisms by Integrating Tumor Genomic Markers and Drug Molecular Structures

Practical Content

1. Preprocess drug molecular structure information

2. Calculate Drug Similarity

3. Build self-attention SWnet on different datasets

4. Evaluation of self-attention SWnet

5. Construction of Multi-task SWnet

6. Constructing a Single-layer SWnet

7. Construction of SWnet with Weighted Layers

Partial Model Case Images

Machine Learning Bioomics

Day One

1. Review of Basic Microbiology Knowledge

2. Introduction to Basic Concepts of Machine Learning

a. What is Machine Learning

b. Supervised learning, unsupervised learning

c. Introduction to Commonly Used Machine Learning Models

3.Confusion Matrix

4.ROC Curve

The Next Day

Introduction to R Language and Practical Operations

1. Overview of R Language

2. R Studio Software and R Package Installation

3. R Language Syntax and Data Types

4. Conditional Statements and Loops

Linux Practical Operations

1. Linux Operating System

2. Installation and Configuration of the Linux Operating System

3. Network Configuration and Service Process Management

4. Remote Login Management of Linux

5. Commonly Used Linux Commands

6. Obtain Genetic Data Under Linux

7. Shell Script and Vim Editor ·

Day Three

Commonly Used Analysis Methods for Microbiome (Practical Operations)

1. Microbial Abundance Analysis

2. Transcriptome Abundance Analysis

3. Phylogenetic Tree Analysis

4. Dimensionality Reduction Analysis

Day 4

Application Case Sharing of Machine Learning in Microbiomics

1. Disease Prediction Application: Using Machine Learning to Predict Disease States Based on Microbiome Data

2. Gut Microbiota Research: Machine Learning Studies the Impact of Diet on Gut Microbes

Day 5

Machine Learning Model Training and Analysis (Hands-on)

1. Data Loading and Data Normalization

2. Build Training Models (GLM, RF, SVM)

3. Model Parameter Optimization

4. Model Error Rate Curve Plotting

5. Confusion Matrix Calculation

6. Key Feature Selection

7. Model validation, ROC curve plotting using the model for prediction

Predicting Host Phenotypes Using Machine Learning Based on Microbiome Data

1. Load Data

2. Data Normalization

3. OUT Feature Processing

4. Construction of Machine Learning Models (Various machine learning methods such as RF, KNN, SVM, Lasso)

5. Plot ROC Curve to Compare Performance Evaluation of Different Machine Learning Models

Predicting Disease Risk Using Machine Learning Based on Clinical Features and Gut Microbiota

1. Load Data

2. Construction of Machine Learning Models (RF, GBM, SVM, etc.)

3. Cross-validation

4. Model Performance Evaluation

Partial Model Case Images

Machine Learning Metabolomics

Content can be scrolled down

Day One

A1 Metabolites and the Development and Application of Metabolomics

(1) Metabolism and Physiological Processes;

(2) Metabolism and Disease;

(3) Untargeted and Targeted Metabolomics;

(4) Spatial Metabolomics and Mass Spectrometry Imaging (MSI);

(5) Metabolomics and drugs and biomarkers;

(6) Metabolic Flux and Mechanism Research.

A2 Metabolic Pathways and Metabolism Database

(1) Introduction to Several Classical Metabolic Pathways;

(2) Three major common metabolite databases: HMDB, METLIN, and KEGG;

(3) Metabolomics Raw Database: Metabolomics Workbench and Metabolights.

A3 Recommended Reference Materials

A4 Metabolomics Experimental Workflow Overview

A5 Chromatography, Mass Spectrometry Hardware and Principle Analysis

(1) Principle and Structure of Chromatographic Analysis;

(2) Selection of chromatograph and chromatographic column;

(3) Chromatographic mobile phase: Gradient elution method;

（4） Ion Source, Quality Analyzer and Quality Detector Analysis;

(5) Principles of Mass Spectrometry and Animation Demonstration;

(6) Liquid Chromatography-Mass Spectrometry (LC-MS);

The Next Day

B1 Metabolite Sample Processing and Extraction

(1) Extraction procedures and precautions for various tissues, blood, and body fluid samples;

(2) Metabolite Extraction Process and Precautions;

(3) Issues related to the transportation and storage of samples and metabolites;

B2 LC-MS Data Quality Control and Database Search

(1) The method for setting QC and Blank samples during the LC-MS experiment;

(2) Data quality control monitoring and analysis during the LC-MS operation process;

(3) Principles of Upstream Metabolomics Analysis —— Based on Compound Discoverer and Xcms Software;

(4) Data conversion, peak extraction, peak alignment, and library search using Xcms software;

B3 R Software Foundation

(1) Installation of R and Rstudio;

(2) Configuration of Rstudio Interface;

(3) Basic Operations and Statistical Calculations in R;

(4) Packages in R: The use of packages, functions, and parameters;

(5) R language syntax, data types and data structures;

(6) R Basic Plotting;

B4 ggplot2

(1) Introduction to ggplot2

(2) The Plotting Philosophy of ggplot2;

(3) The color system of ggplot2;

(4) Practical Data Mining and Plotting with ggplot2;

Day 3

Machine Learning

C1 Application of Supervised Machine Learning in Metabolomics Data Processing

(1) The relationship among artificial intelligence, machine learning, and deep learning;

(2) Regression Algorithms: Starting from Linear Regression, Logistic Regression, and Cox Regression;

(3) PLS-DA Algorithm: Is there still hope for data that shows no difference after PCA dimensionality reduction?

(4) The Meaning and Selection of VIP Score;

(5) Classification Algorithms: Decision Trees, Random Forests, and Bayesian Network Models;

C2 Classification Algorithm Implementation for a Set of Metabolomics Data in R

(1) Data Interpretation;

(2) Drills and Operations;

C3 Application of Unsupervised Machine Learning in Metabolomics Data Processing

(1) Dimensionality reduction in big data processing;

(2) PCA analysis plotting;

(3) Three Common Types of Cluster Analysis: K-means, Hierarchical Analysis, and SOM

(4) R language implementation of heatmaps and hcluster plots;

C4 A R Walkthrough for Dimensionality Reduction and Clustering Analysis of a Set of Metabolomics Data

(1) Data parsing;

(2) Drills and Operations;

Day Four

D1 Online Metabolomics Analysis Webpage Metaboanalyst Operation

(1) Use R to clean the data into the format required by the webpage;

(2) Data format issues for independent groups, paired groups, and multiple groups;

(3) Upstream analysis in Metaboanalyst (raw data peak extraction, peak alignment and library search) (4) Metaboanalyst pipeline, parameter settings and considerations;

(5) Viewing and exporting results from MetaboAnalyst;

(6) Data editing in Metaboanalyst;

(7) Full-process drills and operations.

D2 Metabolomics Data Cleaning and Advanced R Programming

(1) t, fold-change, and response values in metabolomics;

(2) Data cleaning process;

(3) R language tidyverse;

(4) Data Preprocessing: Data Filtering and Data Standardization (Normalization of Samples and Scaling of Metabolites);

(5) Metabolomics Data Cleaning Exercise;

Day 5

E1 Reproduction of Literature Data Analysis Section (1 paper)

(1) In-depth interpretation of literature;

(2) Practical Operation: From Raw Data Download to Image Reproduction;

(3) Students' hands-on practice.

E2 Machine Learning and Metabolomics Top Journal Interpretation (3 articles);

(1) Signal Transduction and Targeted Therapy: A Mouse Brain Tissue Metabolic Atlas Study on the Effects of Starvation on Metabolomics in Different Brain Regions.(Database Type)

(2) A metabolomics analysis of pregnant women's blood throughout pregnancy, published in Cell, identified metabolic biomarkers for predicting gestational age and delivery date. (Biomarker type)

(3) A Nature article on the metabolomics analysis of gut microbiota in pancreatic cancer patients identifies metabolites that can enhance chemotherapy efficacy. (Mechanism study)

Partial Model Case Images

Figure 4.Linking Drug and Protein Information with Attention BlocksAttentionDTAModel

CRISPR-Cas9 Gene Editing

Content can be scrolled down

Day One

Introduction to Gene Editing

1. Introduction to the Basic Concepts of Gene Editing

2. Using Single-Gene Hereditary Disease Database (Practical Operation)

3. Incidence and Corresponding Genes of 5524 Monogenic Hereditary Diseases

4. Use the website to find bases around the mutation site and select gene editing tools (practical operation).

5. History of Gene Editing

6.TALEN

7.Zinc finger

8.Base editor

9.Prime editor

The Next Day

1. How to Choose the Right Type of Cas9 Protein

2. Single Base Editor (Base Editor) gRNA Design and Software (Hands-on)

3.sgRNA Modification

4. Eight Key Points for Manually Designing PegRNA (Practical Operation)

5. Seven Types of PegRNA-Assisted Design Software

6. Find the base sequence of Cas9 with different PAMs

7. Design of Cas9 mRNA and DNA Required for In Vitro Transcription

8. How to Design Primers for Deep Sequencing (NSG)

9. How to Prepare for Deep Sequencing (NGS) Required Samples

10. Sleeping beauty

11. PiggyBac

12. Peptide Nucleic Acids

13. Exon Skipping

14. Introduction to CRISPR Knockout Library (GeCKO v2.0)

15. A Brief Introduction to the Application of Base Editing in Microorganisms

Day Three

1. AAV Delivery (Tissue Targeting)

2. Liposome Delivery

3. Ribonucleoprotein Delivery

4. Polymer Delivery

5.Viral-like particles delivery

6. Exosome Delivery

7. Inorganic Nanoparticle Delivery

8. Electroporation

9. Ultrasound

10. Microinjection

11. Six Articles on PAM High-Throughput Screening

12. Cas9-DNA-gRNA Crystal Structure

13. Monogenic Inherited Disease Cell Customization

14. Customization of Animal Models for Monogenic Hereditary Diseases

Day Four

1.Animal Model

2. Plasmid

3. Fundamentals of Molecular Cloning

4.AAV Design (Practical Operation)

5.Base Editing Article Analysis

6. Overview of 140 Articles Related to Prime Editing

7. How to Improve Prime Editing Efficiency

8.Overview of 87 Research Articles Related to Prime Editing

9. How to Improve Prime Editing Efficiency

10.NGS Primer Design

11. NGS Sequencing Result Analysis

Day Five

1. Gene Editing and Approved Drugs

2. Clinical Trials

3. Major Companies, Scientists, and Patents

4. Side Effects and Withdrawn Products

5.FDA Policy

6. Application of CRISPR in Diagnostics

7.CRISPR library

8. CRISPR and Single-cell Sequencing

9. CRISPR and Epigenetics

10. Application of CIRPSR in Botany

11. Applications of CRISPR in Microbiology

12. Types and Principles of NGS

13. Database of 3 Gene Sequences (Including Plant Genes)

14. Hot Topics in the BE and PE Fields in 2023

15. Design Topics and Evaluation (Practical Operations))

---------------------------------------------------------------------------------------------------------

Single-cell Sequencing and Spatial Multi-omics

Content can be scrolled down

Day One

Bioinformatics Basic Trainingbasics of bioinforma,c

• What is Single-Cell Sequencing Technology (with10X For example)? The application of single-cell sequencing technology.

• Command line Basic Knowledge: What iscommand line?

• Command line Basic Syntax

• Shell scripts

• How to in ChinaHPC (high performance cluster)Submit Homework Online

The Next Day

R Preparation of Basic Knowledge and Single-Cell Data Analysis Tools

• R and RStudio Installation

• Commonly UsedR Package Installation

• Single-cell RelatedR AndRStudio Environment preparation and installation of corresponding packages

• Seurat, ggplot2 Introduction to the Syntax of Commonly Used Packages

Day 3

Basic Theoretical Knowledge and Quality Control Analysis of Single Cells

• Single-cell Experimental Design

• Single-cell Data Generation

• Cellranger Usage and Parsing of Result Files

• Reading and Writing of Single-Cell Data, Quality Spectrum Analysis

• Selection and Quality Control Analysis of Single-Cell Data Quality Control Parameters

• Quality Atlas and Result Analysis of Single-Cell Data After Quality Control

Day 4

Downstream Analysis of Single-Cell DataI

• Standardization Process and Dimensionality Reduction Analysis of Single-Cell Data

• Single-cell dataScMransforma5on Analysis

• Cluster Analysis of Standardized Single-Cell Data

• Doublet Testing

• Cell Annotation

Day Five

Downstream Analysis of Single-Cell DataII

• Analysis of Gene Expression Differences Between Samples

• Enrichment Analysis

• Marker genes Identification

• Integrated Analysis of Data from Different Batches

• Case Analysis

• Other downstream data analysis(It depends on the time.)

Learning Objectives

Deep Learning Protein Design

This course aims to provide students with comprehensive knowledge in the fields of deep learning and protein design. By teaching the fundamental concepts and cutting-edge technologies of deep learning, students will understand its specific applications in bioinformatics, particularly in protein design. Students will learn how to use the mainstream deep learning framework PyTorch for model construction and optimization, and through hands-on practice, they will master key techniques such as protein structure prediction, protein function prediction, and molecular docking. The course will introduce advanced models like AlphaFold and explore their significance in drug discovery. Additionally, through specialized topics such as peptide design and the reverse central dogma, students will gain a comprehensive understanding of strategies for deriving structure from function and designing proteins from scratch.

CADD Computer-Aided Drug Design

Mastering including PDB database, target protein, protein-ligand, protein-ligand small molecule, protein-ligand structure, introduction and usage of notepad, molecular docking, protein-ligand docking, virtual screening, protein-protein docking, protein-polysaccharide molecular docking, protein-hydration docking, Linux installation, gromacs molecular dynamics full practical operation, solvated molecular dynamics simulation

AIDD Artificial Intelligence Drug Discovery and Design

This course enables students to understand the frontier background of drug discovery, learn various common algorithms in the field of artificial intelligence, become familiar with the installation and use of toolkits, master certain algorithm programming skills, and be able to apply computational methods to study drug-related issues. Through extensive case studies and hands-on practice, students will gain the ability to construct AIDD models and perform data analysis.

Deep Learning Genomics

Deeply study and understand the basic framework and logic of deep learning, while mastering the use of fundamental bioinformatics software (such as Linux, R, Python, etc.), enabling learners to better handle genomic data and uncover new knowledge beyond existing information. Constructing well-developed deep learning models can help explore new research ideas and identify potential biological mechanisms, better serving one’s own scientific research and exploration process.

Machine Learning Metabolomics

Familiar with metabolomics and machine learning-related hardware and software; familiar with the entire process of metabolomics from sample processing to data analysis; able to reproduce the figures of at least one CNS or sub-journal level metabolomics article.

Machine Learning Microbiomics

Through the systematic explanation of multiple cases in this training, the participants will learn the machine learning process in microbiome data analysis and be able to quickly apply it to their own research projects and topics.

CRISPR-Cas9 Gene Editing

The course starts from a global perspective and goes from shallow to deep. Through the method of basic introduction + practical application case exercises, it covers the initial principle explanation to the final application practice. After completing this course, you will master the relevant principles and applications of gene editing technology. In addition, you can learn optimization strategies for gene editing systems and how to operate commonly used biological software, which can be quickly applied to your own research projects and topics.

Single-cell Sequencing and Spatial Multi-omics

This course focuses on single-cell sequencing technology, using 10x as an example, to explain the principles and applications of single-cell sequencing technology, the generation of this technology, data quality control, and analysis. Through case studies, participants will gain an in-depth understanding of basic bioinformatics tools, interpretation and analysis of conventional bioinformatics data formats, and the analysis and visualization of single-cell sequencing data. After completing this course, learners will be able to independently handle any type of single-cell data and apply the results to publish academic papers and guide practical clinical research.

Introduction of the Lecturer

Deep Learning Protein Design

The instructor is from Tsinghua University, focusing on computational biology and bioinformatics research, primarily concentrating on the application of deep learning methods (especially large models) in biomedicine. In addition to research work at Tsinghua University, the instructor has conducted research on deep learning prediction of protein-DNA binding at Stanford University and participated in large model-related research on protein-ligand interactions at Microsoft Research. The instructor has published several research papers as the first author/co-first author in SCI journals and served as a reviewer for internationally renowned journals such as PLOS Computational Biology.

CADD Computer-Aided Drug Design

The lecturers are from universities and institutions in China, such as the Chinese Academy of Sciences. They specialize in research areas including deep learning, machine learning, virtual drug screening, computer-aided drug design, AI-driven drug discovery, molecular docking, and molecular dynamics.

AIDD Artificial Intelligence Drug Discovery and Design

The instructor, Mr. Yu, has over a decade of experience in computer algorithm research and programming. His research areas include bioinformatics, deep learning, drug target identification, and adverse drug reactions. He has participated in two projects funded by the National Natural Science Foundation of China and led three provincial-level scientific research projects. As the first author, he has published several SCI papers in well-known journals such as BMC Bioinformatics, Journal of Biomedical Informatics, and International Journal of Molecular Sciences.

Deep Learning Genomics

The main lecturer, Teacher Liu, is a Principal Investigator (PI) in bioinformatics, with over a decade of experience in sequencing data analysis. His research areas include artificial intelligence, natural language processing, functional genomics, transcriptomics, miRNA and target gene network analysis, single-cell sequencing data analysis, time-series analysis of gene regulatory networks, protein-protein interaction network analysis, and multi-omics joint analysis. He has led four projects, including the Provincial Natural Science Foundation, and published 23 SCI papers and one monograph.

Machine Learning Metabolomics

The lecturer is a neuroscience Ph.D. from a 985 university, primarily utilizing technologies such as metabolomics, transcriptomics, and molecular biology to research the pathogenesis and biomarkers of chronic neurological diseases. Skilled in conducting comprehensive non-targeted and targeted metabolomics research using liquid chromatography-mass spectrometry (LC-MS) technology, from sample preparation to data analysis, as well as bioinformatics integration analysis of multi-omics big data. Within the past five years, publications have appeared in J Clin Invest,EBioMedicine, Cell Death Dis, Cell Death Discov, Nanotoxicology and other journals have published 10 SCI papers.

Machine Learning Microbiomics

The main speaker is a core executive from a biotech company, having previously worked at institutions such as MIT and UCSF, with research published in journals like Cell and PNAS over the past five years.

CRISPR-Cas9 Gene Editing Technology

The lecturer holds a Ph.D. in Biomedical Engineering from the University of California, and has conducted research on gene editing at MIT and Harvard University, as well as worked on gene delivery at Yale University. Their articles have been published in journals such as Nature Biomedical Engineering and Nature Communications. They have also worked at a venture capital firm, primarily investing in startups focused on gene editing, single-cell sequencing, and AI-driven drug discovery.

Single-cell Sequencing and Spatial Multi-omics

The lecturer, Dr. Chen, holds a Ph.D. in Bioinformatics from the University of California, Davis, and completed his postdoctoral research in Bioinformatics at the University of California, San Francisco. He currently works in bioinformatics analysis at a well-known biotech company. With years of experience in whole microbial genome analysis, microbial genetic disease resistance analysis, tumor genetic variation analysis (somatic mutations and germline mutations), single-cell and spatial transcriptomics data analysis, he has also developed methods for detecting CNVs in tumor samples. He has published articles as the first or co-author in renowned journals such as Genome Biology and Cell.

Time and Location of the Lecture

Deep Learning Protein Design

2024.07.13-2024.07.14All-day teaching (morning9：00-11：30Afternoon13：30-17：00）2024.07.15-2024.07.16Evening Classes (Evening19：00—Night22：00）2024,07.20-2024.07.20-2024.07.21All-day teaching (morning9：00-11：30Afternoon13：30-17：00）

Tencent Meeting Live Streaming Format

CADD Computer-Aided Drug Design

2024.07.20-2024.07.21All-day teaching (morning9：00-11：30Afternoon13：30-17：00）2024.07.23-2024.07.26Evening Classes (Evening19：00—Evening22：00）

2024.07.27-2024.07.28All-day teaching (morning9：00-11：30Afternoon13：30-17：00）

2024.07.29-2024.07.30Evening Classes (19：00—Evening22：00）

Tencent Meeting Live Streaming Format

AIDD Artificial Intelligence Drug Discovery and Design

2024.07.13-2024.07.14All-day teaching (morning9：00-11：30Afternoon13：30-17：00）2024.07.15-2024.07.16Evening Class (Evening19：00—Evening22：00）2024,07.20-2024.07.20-2024.07.21All-day teaching (morning9：00-11：30Afternoon13：30-17：00）