Structural Bioinformatics and Molecular Modeling

Module: Structural Bioinformatics
1. Introduction to bioinformatics
2. Genome organisation and evolution
3. Databases, archives, and information retrieval
4. Substitution matrices, pairwise and multiple alignments, database searching, and phylogenetic trees
5. Protein structure and architecture
6. Protein structure prediction and validation: comparative modelling, threading, and ab initio approaches
7. Introduction to systems biology
8. Computer lab sessions (1 ECTS)
Module: Molecular Modelling: Basic Methodologies
1. Introduction to quantum mechanics
o The Schrödinger equation and the wave function
o The Hamiltonian operator
o Main approximations in QM calculations: Born-Oppenheimer, Hartree-Fock, LCAO
o Basis sets and the SCF cycle
o Potential energy surface and geometry optimisation
o QM methods for the calculation of molecular properties: ab initio, semi-empirical, and DFT
2. Introduction to molecular mechanics
o Force fields
o Solvent models and periodic boundary conditions
o Geometry optimisation
3. Conformational search
o Systematic methods
o Stochastic methods
4. Molecular dynamics (MD)
o Equations of motion and trajectory calculation
o Microcanonical (NVE), canonical (NVT), and isothermal-isobaric (NPT) ensembles
o Trajectory analysis: energy profiles, RMSD, RMSF, geometric parameters, hydrogen bonds, cluster analysis, principal component analysis
o Applications and limitations of MD
5. Enhanced sampling techniques
o Elements of statistical thermodynamics
o Simulated annealing
o Umbrella sampling
o Replica exchange MD
o Metadynamics
o Accelerated MD
6. Free energy calculations in complex systems
o Potential of mean force (PMF)
o Alchemical perturbations: Free Energy Perturbation and Thermodynamic Integration
o End-point methods: MM-PBSA
Module: Computational Methodologies in Biopharmaceutical Development
1. Introduction to cheminformatics and computer-aided drug design (CADD)
o Fundamentals of computer science
o Ligand-based and structure-based approaches
o Introduction to QSAR modelling and virtual screening
2. Representation of small molecules in silico
o Molecular graphs, matrices, and connection tables
o CT files, line notations, and molecular fingerprints
3. Molecular descriptors
o Definition, characteristics, and classification by dimensionality
o ECFP, LogP, geometric descriptors, electronic descriptors, molecular fields, MLP, Virtual LogP
o Advantages and limitations of the main descriptor families
4. History of QSAR modelling
5. Traditional regression QSAR
o Meaning and requirements
o Steps in developing a regression predictive model
o Least squares algorithm and multiple linear regression
o Principal component analysis (PCA) and genetic algorithms
o R², Pearson's R, PRESS, Q²
o Model application and applicability domain analysis
6. Modern classification QSAR
o Meaning and requirements
o Introduction to machine learning concepts and the Random Forest algorithm
o Steps in developing a classification predictive model
o Precision, Recall, Matthews Correlation Coefficient (MCC)
o Model application and applicability domain analysis
7. Theory of molecular docking simulations
o Meaning, classification, and model validation
o Searching algorithms
o Scoring functions: meaning and types
8. Theory of virtual screening
o Ligand-based and structure-based strategies
o Steps in developing a predictive model and the concept of enrichment
o Sensitivity, specificity, accuracy, top-N% enrichment factor, ROC curve.