Computational Drug Discovery & Intensive Genomics

Aim

Workshop Objectives

Workshop Structure

Day 1: Genomics Data Processing & ML Foundations

• Retrieval and preprocessing of genomic datasets from NCBI GEO / ENA
• (FASTQ/FASTA formats)
• Quality control and trimming using FastQC and preprocessing pipelines
• Feature extraction from sequences (k-mers, GC content, embeddings)
• Encoding biological sequences using one-hot, k-mer vectors, and transformer
• embeddings
• Dimensionality reduction using PCA / t-SNE / UMAP on genomic features
• Classification of gene sequences using SVM / Random Forest models
• Variant analysis and SNP classification using ML pipelines
• Building a basic genomic prediction model (disease vs normal classification)

Day 2: Machine Learning in Drug Discovery

• Retrieval and curation of chemical datasets from PubChem / ChEMBL
• Molecular descriptor calculation using RDKit (physicochemical properties)
• Conversion of molecules into fingerprints (ECFP, MACCS)
• QSAR modeling using regression/classification ML models
• Drug-target interaction prediction using matrix factorization / ML models
• Training ML models for binding affinity prediction (regression tasks)
• Virtual screening using ML-based filtering of compound libraries
• Model evaluation using ROC-AUC, precision-recall, RMSE metrics

Day 3: Integrated AI Pipelines & Advanced Applications

• Integration of genomic & chemical data for target identification
• Multi-omics data fusion using ML-based feature integration techniques
• Building a drug response prediction model using genomic signatures
• Implementation of deep learning models (CNNs for sequence analysis)
• Training a Graph Neural Network (GNN) for molecular property prediction
• Explainability using SHAP / feature importance in biological models
• End-to-end pipeline: target identification → lead prediction → validation
• Deployment of ML models using Streamlit / API for real-time predictions

Important Dates

Workshop Outcomes

Fee Structure

What You’ll Gain