Mandatory Module M3: Computational Biosystems Science and Engineering

Note: Draft version

Coordination: Bruce Tidor (MIT) + Eugénio Ferreira (U Minho)

This course provides an introduction to computational biology, emphasizing the fundamentals of nucleic acid and protein sequence and structural analysis. It also includes an introduction to the analysis of complex biological principles. Covers principles and methods used for sequence alignment, motif finding, structural modeling, structure prediction, and network modeling. This course is based on a multi-disciplinary approach for obtaining, modeling, organizing and managing large volumes of data, obtained experimentally or computationally. The central objective is to educate students in the techniques required to carry out research in this area.

Topics

Lecture 1 (2h, Eugénio Ferreira) : Introductory Topics

Course Overview, Learning outcomes
Introduction to Bioinformatics, Computational Biology and Systems Biology
Introduction to format and pace of classes. Assignment of papers for discussion and home work.

Module I - Sequence and Structural Bioinformatics

Lecture 1 (2 h; Ana Teresa Freitas): Sequence comparison algorithms; Edit distance and alignments; Global and local sequence alignment; Multiple alignment

Lecture 2 (1.5 h; Ana Teresa Freitas): Motif finding and Motif inference: Combinatorial pattern matching; Probabilistic algorithms

Lecture 3 (2h; Cláudio Soares & António Baptista): Principles of molecular mechanics/dynamics. Application to biomolecules (1h; Cláudio Soares); Biomolecular electrostatics using continuum models (1h; António Baptista)

Module II - Differential Equation Modeling (Bruce Tidor)

Lecture 1 (1.5 h): Overview and Introduction to Network Modeling with Ordinary Differential Equations.
Overview of Model Hierarchy (from statistical data mining to mechanistic modeling); Role of Modeling (encapsulate knowledge; make predictions); ODE Model Formulation; Transient and Steady-State Behavior

Lecture 2 (1.5 h): Advanced Network Modeling with Ordinary Differential Equations
ODE Model Simulation; ODE Model Optimization; Hybrid Models and Time-locked Behavior

Lecture 3 (2 h): Robustness of Network Models and Stochastic Simulation
Discussion of Barkai & Leibler (“Robustness in Simple Biochemical Networks,” Nature 387: 913–917, 1997); Gillespie Algorithm

Lecture 4 (1.5 h): Nonlinear Dynamics
Phase Portraits; Stationary Points and Their Character; Linearization; A Simple Genetic Switch

Module III – Analysis of High-Throughput Data

Lecture 1 (2 h; André Valente): Transcriptome Analysis. Introduction to microarray technology …

Lecture 2 (1.5 h; Pedro Santos): Expression Proteomics: experimental approaches and applications
Theory of 2D gel electrophoresis; Quantitative proteomics, including stable isotope labelling and DIGE; Experimental design for proteomic studies; Mass spectrometry for protein analysis and identification; Analysis of post-translational modifications; Examples of application of expression proteomics in studies focusing microorganism-environment interactions.

Lecture 3 (1 h; Miguel Teixeira): Functional Genomics
Functional genomics: experimental approaches in the EUROFAN project. The case study of the functional analysis of new yeast genes encoding proteins of the Major Facilitator Superfamily (MFS), involved in Multiple Drug Resistance (MDR) in yeast.

Lecture 4 (1 h; Arlindo Oliveira & Ana Teresa Freitas): Data mining & Analysis of High-Throughput Data
Microarray data analysis using unsupervised and supervised methods. Clustering and biclustering. Decision trees.

Lecture 5 (2h; Rui Oliveira): Fluxome and metabolome analysis
Metabolic pathway analysis - Elementary Modes. Metabolic Flux Analysis. Flux Balance Analysis

Demonstration (2h): Expression proteomics: data analysis of yeast proteome alterations induced by chemical stress. Comparative analysis of 2D gel using Image Master Platinum software (Amersham Biosciences)
Genome-wide expression analysis of the yeast response to chemical stress using bioinformatics tools: GO-based grouping and metabolic pathway assignment; Prediction of the transcriptional regulatory networks underlying the observed expression changes, using the YEASTRACT database; Identification of over-represented motifs in the promoter regions of co-regulated genes, using the YEASTRACT computational tools

Module IV - Biomolecular Kinetics

Lecture 1 (1.5 h; Dane Wittrup): Kinetics and Equilibria of Noncovalent Interactions
Molecular recognition fundamentals. Monovalent equilibrium relationships. Multivalency – cooperativity and avidity.
Association and dissociation kinetics: pseudo-first order. Environmental effects (temperature, pH, ionic strength).

Lecture 2 (1.5 h; Dane Wittrup): Theory & Practice of Biomolecular Measurements
Fluorescence theory. Fluorophores & intrinsically fluorescent proteins. Intrinsic and localization methods of detection of molecular complexes. General issues in experimental design. Nonlinear regression

Lecture 3 (1.5 h; Dane Wittrup): Protein Synthesis and Cellular Localization dynamics
Biomolecular dynamic material balances. Gene expression models. Protein trafficking – secretion, endocytosis, localization

Lecture 4 (1.5 h; Dane Wittrup): Reaction & Diffusion
Spatial heterogeneity: compartmental models vs. transport. Scaling analyses. Shrinking core model. Tumor targeting applications

Laboratory: Student presentations of analyses of research papers: Teams of 4; 30 minute presentations.

Module V – Integrative Seminars

Lecture 1 (1 h; Isabel Rocha): Genome-scale models and Metabolic Engineering
Re-construction of metabolic networks from sequence data. Flux Balance Analysis. Applications of genome-scale models: in silico metabolic engineering

Lecture 2 (1 h; Miguel Rocha): Alternative modelling schemes (Petri Nets, ANN)
Overview of qualitative and quantitative models for metabolic, regulatory and signalling networks. Petri Nets and other graph-based approaches for the representation and simulation of discrete and continuous models. Artificial Neural networks and other non-linear models: training and simulation. Algorithms for data-driven model optimization: traditional methods, nature-based optimization metaheuristics.

Lecture 3 (1 h; João Aires-Sousa): Chemoinformatics. Representation of chemical compounds. Standard structure exchange formats. Representation of chemical reactions and databases of metabolic reactions. Structural descriptors for quantitative structure-activity relationships (QSAR).

Lecture 4 (1 h; José Cardoso Menezes): Chemometrics. Exploratory Data Analysis of Large and Complex Data Structures. Multivariate Linear and Non-Linear Data Classification. Multivariate Linear and Non-Linear Data-Based Modelling and Calibration.