AI-Powered Metabolomics & Precision Medicine: Disease Biomarker Discovery, Multi-Omics Analysis & Machine Learning

Aim

Workshop Objectives

Workshop Structure

📅 Day 1: Metabolomics Data Analysis & Disease-Associated Metabolic Signatures

Focus: Process metabolomics datasets and identify metabolic differences associated with disease or biological conditions.

Topics Covered

• Introduction to metabolomics and its role in precision medicine
• Targeted vs untargeted metabolomics
• Metabolite data formats and experimental design
• Data quality control, missing-value handling, normalization, and transformation
• Exploratory data analysis and metabolite distributions
• Univariate statistical analysis and differential metabolite analysis
• Multivariate analysis and dimensionality reduction
• PCA and clustering for metabolic-pattern discovery
• Heatmaps, volcano plots, and metabolite-profile visualization
• Metabolite annotation and biological interpretation

🛠️ Hands-on

Analyse a curated metabolomics dataset, preprocess the data, identify significantly altered metabolites, and visualize disease-associated metabolic patterns.

Workflow:
Metabolomics Data → QC → Normalization → Statistical Analysis → PCA/Clustering → Metabolic Signatures

🧰 Tools Covered

Python, R, pandas, NumPy, scikit-learn, MetaboAnalyst, matplotlib, ggplot2, Google Colab

📅 Day 2: Biomarker Discovery & Metabolic Pathway Analysis

Focus: Identify candidate biomarkers and interpret their biological relevance through pathway-level analysis.

Topics Covered

• Principles of metabolomics-based biomarker discovery
• Selection and prioritization of candidate metabolites
• Biomarker performance and diagnostic classification concepts
• ROC curves, AUC, sensitivity, and specificity
• Feature selection and dimensionality reduction
• Metabolite correlation and metabolic-network analysis
• Pathway enrichment and metabolite-set enrichment
• Metabolic pathway interpretation
• Disease-associated metabolic signatures
• Integration of metabolite and biological pathway information

🛠️ Hands-on

Prioritize candidate biomarkers, evaluate their discriminatory performance, perform pathway analysis, and construct a disease-associated metabolic signature.

Workflow:
Differential Metabolites → Feature Selection → Biomarker Evaluation → Pathway Analysis → Biological Interpretation

🧰 Tools Covered

MetaboAnalyst, Python, R, scikit-learn, pandas, Plotly, pathway/metabolite databases, Google Colab

📅 Day 3: Multi-Omics Integration & AI-Based Precision Medicine

Focus: Integrate metabolomics with other molecular datasets and develop interpretable ML models for disease prediction.

Topics Covered

• Principles of multi-omics integration
• Integration of metabolomics with genomics and transcriptomics
• Correlation-based and network-based multi-omics analysis
• Identification of cross-omics molecular signatures
• Machine-learning approaches for disease classification
• Random Forest and XGBoost-based prediction
• Feature engineering and feature selection
• Model training, validation, and cross-validation
• Evaluation using accuracy, precision, recall, F1-score, ROC-AUC, and related metrics
• Explainable AI and feature-importance analysis
• SHAP-based interpretation of metabolic biomarkers
• Biomarker-panel development and precision-medicine applications

🛠️ Hands-on

Integrate multi-omics datasets, train an ML disease-classification model, identify important metabolites and molecular features, and develop an interpretable biomarker signature.

Workflow:
Multi-Omics Data → Integration → Feature Engineering → ML Prediction → XAI → Biomarker Panel → Precision-Medicine Insights

🧰 Tools Covered

Python, R, pandas, scikit-learn, XGBoost, SHAP, Cytoscape, Plotly, Google Colab

Who Should Enrol?

Important Dates

Workshop Outcomes

Fee Structure