Topic One

(Live)

CADD Protein Structure Analysis, Virtual Screening, Molecular Docking

(Protein-Protein, Protein-Peptide)

Topic Two

(Live)

AMBER Molecular Dynamics Energy Optimization and Analysis, Binding Free Energy Calculation Topic

Topic Three

(Live)

AIDD Artificial Intelligence (Shallow Learning and Deep Learning) Drug Discovery

Time

Course Name

Course Content

Day One

Morning

Basics of Biomolecular Interactions

1. Methods for studying biomolecular interactions

1.1 Principles of protein-small molecule and protein-protein interactions

1.2 Molecular docking to study biomolecular interactions

1.3 Protein-protein docking to study molecular interactions

Protein Database

1. Introduction to the PDB database

1.1 Functions of the PDB protein database

1.2 Resources for obtaining PDB protein data

1.3 Importance of the PDB protein database for drug development

2. Using the PDB database

2.1 Types of target protein structures, data interpretation and download

2.2 Downloading target protein structure sequences

2.3 Background analysis of target proteins

2.4 Ways to obtain related data resources

2.4 Batch download of protein crystal structures

Day One

Afternoon

Protein Structure Analysis

1. Introduction to Pymol software

1.1 Software installation and initial settings 1.2 Basic knowledge introduction (such as hydrogen bonds, etc.)

2. Using Pymol software

2.1 Diagrams of protein-small molecule interactions

2.2 Diagrams of protein-protein interactions

2.3 Surface diagrams of proteins and small molecules, electrostatic potential representations

2.4 Superimposing and comparing protein and small molecule structures

2.5 Drawing interaction forces

2.6 Making Pymol animations

Example explanations and exercises:

(1) Interaction of nilotinib with the target, listing the interacting amino acids and exporting the binding mode diagram

(2) Making a binding pocket surface diagram

(3) Superimposing the PDB structure of the Bcr/Abl target

(4) Making a protein interaction animation

(5) For the protein crystal complex of ACE2 and the COVID-19 Spike protein, making a protein-protein interaction

Morning of Day Two

Homology Modeling

1. Introduction to the principles of homology modeling

1.1 Functions and usage scenarios of homology modeling

1.2 Methods of homology modeling

2. Swiss-Model homology modeling;

2.1 Searching for homologous proteins (using methods like blast)

2.2 Protein sequence alignment

2.3 Selecting protein templates

2.4 Building protein models

2.5 Model evaluation (protein Ramachandran plot)

2.6 Protein model optimization

Example explanations and exercises: Modeling using the 2019-nCoV spike protein sequence, evaluating the model based on corresponding parameters and methods

Afternoon of Day Two

Small Molecule Construction

1. Introduction to ChemDraw software

1.1 Constructing small molecule structures

1.2 Calculating the physicochemical properties of small molecules (such as molecular weight, clogP, etc.)

Example explanations and exercises: Constructing macrocyclic compounds, amino acids, DNA, RNA, etc.

Small Molecule Compound Library

1. Small molecule databases

1.1 Introduction and usage of databases like DrugBank, ZINC, ChEMBL

1.2 Introduction and usage of natural product and traditional Chinese medicine databases

Morning of Day Three

Biomolecular Interactions I

1. Basics of molecular docking

1.1 Principles of molecular docking and introduction to docking software

2. Using molecular docking software (Autodock)

2.1 Semi-flexible docking

2.1.1 Preparing small molecule ligands for optimization

2.1.2 Optimizing protein receptors and preparing coordinate files

2.1.3 Calculating grid points for protein receptors

2.1.4 Performing semi-flexible docking calculations

2.2 Evaluating docking results

2.2.1 Comparing crystal structure conformations

2.2.2 Evaluating docking results from an energy perspective

2.2.3 Evaluating docking results through clustering analysis

2.2.4 Selecting the optimal binding conformation

2.2.5 Comparing docking results with known active compounds

Example explanations and exercises: Semi-flexible docking of kinase Bcr/Abl target inhibitors

Afternoon of Day Three

Biomolecular Interactions II

2.3 Flexible docking

2.3.1 Preparing small molecule ligands for optimization

2.3.2 Optimizing protein receptors and preparing coordinate files

2.3.3 Calculating grid points for protein receptors

2.3.4 Performing flexible docking calculations and evaluating results

2.3.6 Comparing and selecting between semi-flexible and flexible docking

Example explanations and exercises: Flexible docking of Bcr/Abl target inhibitors

Virtual Screening

3. Using molecular docking for virtual screening (Autodock)

3.1 Defining virtual screening, building processes, and demonstrations

3.2 Selecting target proteins and obtaining compound libraries

3.3 Conducting virtual screening

3.4 Analyzing results (scoring values, energy, and interaction analysis)

Example explanations and exercises: Virtual screening of Bcr/Abl target inhibitors

Small Molecule Format Conversion

1. Introduction and usage of openbabel

1.1 Introduction to openbabel software

1.2 Types of small molecule structures

1.3 Converting types of small molecule structures

Morning of Day Four

Expanding docking usage scenarios (Part I)

1. Protein-protein macromolecular docking

1.1 Application scenarios of protein-protein docking

1.2 Introduction to related programs

1.3 Preparing receptors and ligands for optimization

1.4 Loading receptor and ligand molecules

1.5 Setting docking sites for protein-protein interactions

1.6 Analyzing and interpreting results from protein-protein docking

Example explanations and exercises: Docking of COVID-19 Spike protein and host protein ACE2

2. Protein-peptide docking

2.1 Introduction to protein-peptide interactions

2.2 Preprocessing protein-peptide molecules

2.3 Docking protein-peptide molecules

2.4 Displaying and analyzing docking results

Example explanations and exercises: Docking of COVID-19 target 3CL with peptide/class-peptide inhibitors

Afternoon of Day Four

Expanding docking usage scenarios (Part II)

4. Small molecule docking with small molecules

4.1 Introduction to small molecule-small molecule interactions

4.2 Preprocessing small molecule structures

4.3 Docking small molecules with small molecules

4.4 Displaying and analyzing docking results

Example explanations and exercises: Docking of cyclodextrin with drug small molecules

5. Nucleic acid-small molecule docking

5.1 Application scenarios of nucleic acid-small molecule interactions

5.2 Introduction to nucleic acid-small molecule interactions

5.3 Preprocessing nucleic acid-small molecules

5.4 Docking nucleic acid-small molecules

5.5 Displaying and analyzing related results

Example explanations and exercises: Docking of DNA G-quadruplex and ligand molecules

Covalent Docking

6. Principles and application scenarios of covalent docking

6.2 Preprocessing proteins and covalently binding ligands

6.3 Covalent docking of drug molecules with target proteins

6.6 Displaying and analyzing related results

Example explanations and exercises: Covalent docking of kinase target EGFR inhibitors

Morning of Day Five

Fragment-Based Drug Design

1. Fragment-based drug design and discovery

1.1 Constructing fragment-based compound libraries 1.2 Skeleton replacement 1.3 Fragment linking 1.4 Fragment growth 1.5 Generating compound libraries based on pharmacophores 1.6 Generating compound libraries based on protein binding pockets 1.7 Generating compound libraries based on molecular descriptors 1.8 Constructing compound libraries based on BREED rules 1.9 Screening compound libraries based on fragments

Example explanations and exercises: Optimization and modification of Bcr/Abl target inhibitors based on fragments

Afternoon of Day Five

Structure-Activity Relationship Analysis

1. Building 3D-QSAR models (using Sybyl software)

1.1 Constructing small molecules

1.2 Creating a small molecule database

1.3 Adding charges and optimizing energy of small molecules

1.4 Determining and superimposing active conformations of molecules

1.5 Creating 3D-QSAR models

1.6 Building CoMFA and CoMSIA models

1.7 Validating models with test sets

1.8 Analyzing model parameters

1.9 Analyzing model isopotential maps

1.10 Guiding drug design with 3D-QSAR models

Example explanations and exercises: Building structure-activity relationship models and predicting activity for kinase target Bcr/Abl inhibitors

All Day on Day Six

Molecular Dynamics Simulation

1. Introduction to molecular dynamics (using GROMACS software)

1.1 Basic principles of molecular dynamics

1.2 Introduction to the Linux system

1.3 Introduction to molecular dynamics software (Gromacs)

2. Conducting molecular dynamics simulations with Gromacs

2.1 Processing ligand molecules

2.2 Processing protein structures

2.3 Modifying protein coordinate files

2.4 Modifying topology files

2.5 Constructing boxes and adding solvents

2.6 Balancing system charge

2.7 Energy minimization

2.8 NVT balancing

2.9 NPT balancing

2.10 Outputting molecular dynamics simulations

3. Analyzing molecular dynamics results

3.1 Observing trajectory files

3.2 Plotting energy data

3.3 Processing trajectory corrections

3.4 Analyzing radius of gyration

3.5 Calculating RMSD changes of protein conformations

3.6 Calculating RMSF changes of atomic positions

3.7 Clustering protein-ligand conformations

3.8 Analyzing hydrogen bond interactions between proteins and ligands

3.9 Analyzing interaction energies between proteins and ligands

Example explanations and exercises:

(1) Simulating pure proteins of lysozyme in water

(2) Simulating complexes of T4 lysozyme and ligands

Q&A

Answering questions related to the last three days of learning

Day One

Time

Course Content

Main Knowledge Points

AM

9:00~9:50

1. Introduction to Molecular Dynamics

Teaching Objective: Understand the content in this direction, theoretical basis, and research significance.

1.1 Basic assumptions of molecular mechanics

1.2 Main forms of molecular mechanics

2 Molecular Force Fields

2.1 Introduction to molecular force fields

2.2 Principles of molecular force fields

2.3 Classification and applications of molecular force fields

AM

10:00~10:50

2. Introduction to LINUX

Teaching Objective: Master the numerical computation platform, familiarize with computer languages, and be able to use the vim editor to edit files simply.

3 Introduction to LINUX

3.1 User groups and permissions3.2 Directory file attributes3.3 Basic LINUX commands3.4 LINUX environment variables3.5 Common shell command exercises

AM

11:00~12:00

PM

14:00~14:50

3. Introduction and Installation of AMBER

Teaching Objective: Understand the historical development of Amber software, familiarize with installation environments, and support environment compilation.

4 Introduction and Installation of AMBER

4.1 Introduction and installation of GCC4.2 Introduction and installation of Open MPI4.3 Running AMBER installation4.4 LINUX operation exercises

PM

15:00~15:50

PM

16:00~17:00

Day Two

Time

Course Content

Main Knowledge Points

AM

9:00~9:50

4. Obtaining Research Object Models

Teaching Objective: How to establish research objects, familiarize with the usage of protein databases, and how to model research objects.

5 Preprocessing Model Files

5.1 Introduction to model sources5.2 Introduction to protein files

AM

10:00~10:50

5. Constructing Research Object Models

Teaching Objective: Familiarize with the preprocessing workflow of models, master the writing of input files, and be able to independently complete the preparatory work before system dynamics.

6 Preprocessing Model Files

6.1 Protein preprocessing6.2 Small molecule preprocessing

AM

11:00~12:00

6.3 Introduction to AMBER force fields

6.4 Introduction to topology files and coordinate files

6.5 Generating top and crd files

6.6 Using the tleap module

Case Practice:

Ø Preprocessing of HIV-1 complex

PM

14:00~14:50

6. Molecular Dynamics Simulation

Teaching Objective: Molecular dynamics workflow, principles of AMBER software dynamics, and completing operational exercises for molecular dynamics simulations.

7.1 Significance and methods of energy optimization

7.2 Significance and methods of adjusting simulation temperature

7.3 Classification and selection of solvent models

7.4 Writing input files for dynamics simulation

7.5 Running molecular dynamics simulations

7.6 Interpreting output content

7.7 Exercises and Q&A

Case Practice:

Ø Energy optimization and molecular dynamics simulation of HIV-1 complex

PM

15:00~15:50

AM

16:00~17:00

Day Three

Time

Course Content

Main Knowledge Points

AM

9:00~9:50

7. Calculating Binding Free Energy

Teaching Objective: Familiarize with the significance of binding free energy calculations, MMPBSA methods, and processes.

8 Enthalpy Change Calculations

8.1 Analyzing and retrieving experimental data8.2 Principles of MM/PBSA binding free energy calculations8.3 Explanation and classification of GB models8.4 Writing input files for enthalpy change8.5 Interpreting enthalpy change results

AM

10:00~10:50

9 Entropy Change Calculations

9.1 Principles of Nmode entropy change calculations9.2 Writing input files for entropy change9.3 Interpreting entropy change results9.4 Comparing experimental and theoretical values

Case Practice:

Ø Binding free energy calculations between HIV-1 and inhibitors

AM

11:00~12:00

PM

14:00~14:50

8. Visualization Software

Teaching Objective: Familiarize with the channels for obtaining visualization software, installation, and basic usage, using visualization software to assist in research work.

10 3D Visualization Analysis

10.1 Installing and using VMD10.2 Installing and using Pymol

PM

15:00~15:50

9. Obtaining Trajectory Features Based on Molecular Dynamics

Teaching Objective: Starting from conformations obtained from dynamics simulations, gain insights into conformational transitions, explain experimental imaginations, and predict experimental results.

11 Conformational Analysis

11.1 RMSD analysis11.2 B-Factory analysis11.3 RMSF analysis

AM

16:00~17:00

Day Four

Time

Course Content

Main Knowledge Points

AM

9:00~9:50

10. Analyzing Interaction Mechanisms Based on Energy

Teaching Objective: From an energy perspective, analyze interaction mechanisms between molecules, residues, and important groups, providing theoretical guidance for experiments.

12 Energy Analysis

12.1 Residue decomposition (interaction analysis)12.2 Alanine scanning (finding hotspot residues)12.3 Hydrogen bond networks (salt bridges, pi-pi conjugation, and other interactions)12.4 Exercises and Q&A

AM

10:00~10:50

AM

11:00~12:00

PM

14:00~14:50

11. Reproducing Classic Works

Teaching Objective: Guide beginners to understand classic works in this direction, reproduce important analysis methods in the work, and deepen students’ understanding of this direction.

13 Reproducing Classic Literature Works (Students are required to download and read carefully before class)

13.1 Nanoscale 2020, 12, 7134−7145.

(a) RMSD and 2D_RMSD testing the stability of simulations(b) Comparing and analyzing binding free energy(c) Thermodynamic integration to calculate relative binding free energy(d) Analyzing hydrogen bond networks(e) Residue decomposition predicting hotspot amino acids

13.2 J. Chem. Inf. Model. 2021, 61, 7, 3529–3542

(a) Alanine scanning predicting hotspot amino acids(b) Residue contact analysis(c) Analyzing temperature factors (B-factor)(d) Analyzing solvent accessible surface area(e) Reproducing dissociation mechanisms through umbrella sampling dynamics

13.3 Exercises and Q&A

PM

15:00~15:50

PM

16:00~17:00

Time

Course Content

Course Content

First Morning

Computer-Aided Drug Molecular Design

1. Introduction to Computer-Aided Drug Design (CADD)

2. Basic methods of CADD

2.1 Molecular Docking

2.2 Pharmacophore

2.3 QSAR and QSPR

2.4 Introduction to various pharmaceutical research databases

First Afternoon

Installation and Configuration of Anaconda3

3. Anaconda3

3.1 Pandas

3.2 NumPy

3.3 RDKit

3.4 scikit-learn

3.5 Pytorch

3.6 Tensorflow

3.7 DeepChem

3.8 XGBoost

Second Morning

Introduction to AIDD

—— Classification and Regression Tasks

1. Building and Applying Classification Models

1.1 Principles of Logistic Regression Algorithm

1.2 Principles of Naive Bayes Algorithm

1.3 Principles of k-Nearest Neighbors Algorithm

1.4 Principles of Support Vector Machine Algorithm

1.5 Principles of Random Forest Algorithm

1.6 Principles of Gradient Boosting Algorithm

2. Introduction to Molecular Features

2.1 Molecular Descriptors

2.2 Molecular Fingerprints

2.3 Molecular Graphs

Second Afternoon

Drug Discovery Based on Shallow Machine Learning (Goal: Guide students to implement toxicity prediction models based on other three algorithms like KNN, SVM, XGBoost, and use them for predicting the toxicity of small molecule compounds)

3. Model Evaluation Methods

3.1 Cross-validation

3.2 External validation

3.3 Common evaluation metrics for classification models

3.4 Confusion matrix

3.5 Accuracy

3.6 Sensitivity

3.7 Specificity

4. Parameter Optimization and Model Selection

4.1 Hyperparameter optimization

4.2 Criteria for model selection

Model example explanations and exercises (literature reproduction):

Ø Using a given dataset as an example, building and using a toxicity prediction model related to CYPs inhibitors based on the random forest algorithm

Third Morning

Drug Discovery Based on Shallow Learning—Regression Tasks

1. Building and Applying Regression Models

2. Common Evaluation Metrics for Regression Models

2.1 MSE

2.2 MAE

2.3 R2

3. Model Selection

3.1 Hyperparameter Optimization

3.2 Selecting the Optimal Model

Third Afternoon

Virtual Screening Based on Shallow Learning Classification (Goal: Guide students to implement pIC50 value prediction models based on other three algorithms and use them for predicting pIC50 values of small molecule compounds)

Model example explanations and exercises (literature reproduction):

Ø Using tubulin as an example, the practical role of machine learning classification tasks in drug discovery

Ø Using a given dataset as an example, building and using a pIC50 value prediction model based on the random forest algorithm

Fourth Day

Drug Discovery Based on Deep Learning

1. Development History of Deep Learning

1.1 Applications of Deep Learning in Drug Development

1.2 DNN

1.3 GCN

1.4 GAT

1.5 KGCN

1.6 FP-GNN

2. Common Frameworks for Deep Neural Networks

2.1 PyTorch

2.2 TensorFlow

3. Introduction and Usage of DEEPCHEM

Model example explanations and exercises (literature reproduction):

Ø Integrated machine learning methods of DEEPCHEM and how to load and use them

Fifth Day

Practical operations using mainstream deep learning models such as DNN, GCN, GAT

Example explanations and exercises (literature reproduction):

Ø Using a given dataset as an example, predicting small molecule anti-breast cancer activity using mainstream deep learning models

Ø Basic ideas, logic, and topic design for breast cancer prediction