Bioinformatics and Functional Genomics

Bioinformatics is an interdisciplinary field that develops methods and software tools for understanding biological data. As an interdisciplinary field of science, bioinformatics combines Computer Science, Biology, Mathematics, and Engineering to analyze and interpret biological data. Bioinformatics has been used for in silico analyses of biological queries using mathematical and statistical techniques.

In the late 1980s as we intend it nowadays.

Skills can be learned/built. The main traits you must have are Passion, Flexibility, Resiliency, Ability to listen.

Understanding the biology underneath the data using an ensemble of skills from many disciplines.

45% bio, 25% informatics, 30% mathematics and statistics

The master will start on March 15th. The venue for lectures and tutorials will be at Istituto Nazionale Genetica Molecolare INGM, Via Francesco Sforza, 35 – 20122 Milano, Italia

Mainly R and Python for analyzing data with their companion libraries for Bioinformatics, Bioconductor and BioPython/NumPy respectively. But those are not the only languages that we are going to use: Javascript, for instance, is used for building HTML pages or interactive charts; Perl is widely used by UCSC tools and Ensembl. Bash is not a programming language per se but sometimes is better scripting that prototyping code in a programming language. Groovy is the main language used to develop NextFlow, so some Groovy may be useful.

• Sequencing technologies, from the beginning to the most advanced solutions

• Introduction to sequencing technologies from a data analysts view

• Applications of sequencing technologies

• NGS Terminolgy

• DNA-seq

• RNA-seq

• ChIP-seq

• ATAC-seq

• Single Cell RNAseq

• Introduction to Downstream Analysis

• Databases for NGS analysis

• Accessing public sequencing repositories

Populations and sets of data

Univariate data (location and variability of the data, empirical distributions, histogram, boxplot; paired data: table, scatterplot, sample correlation coefficient)

Bayes theorem, random variables and their relative distributions (discrete: Bernoulli, Poisson; continuous: Weibull, Gaussian), laws obtained by means of gaussian distribution (t-distribution, chi-square distribution, Fisher distribution), weak law of large numbers, central limit theorem

Point estimation (estimators: bias, mean squared error, efficiency) and interval estimation (confidence intervals for the mean of normal distributions with known and unkown variance, Confidence intervals for the variance of normal distributions confidence interval for the mean of Bernoulli distributions)

Terminology, tests of hypothesis for the mean of normal distributions with known and unknown variance, Test of hypothesis for the variance of normal distribution, for the mean of Bernoulli distributions and for two means and two variances

linear regression models, multiple regression models

PCA, Classification (kNN) and Discriminant Analysis (LDA), Multiple Regression Analysis (PLS)

key concepts and terminology ( graph, node,edge, path, tree, adjacency matrix)

• models and properties (shortest-longest path, diameter, degree or connectivity, random vs scale-free networks,power law degree distributions, clustering coefficient)

• network structure (hubs, bottlenecks, information flow)

relational network, correlation network, regulatory network. WGCNA an example of correlation network

• basic concepts (vectors and matrices, eigenvalues and eigenvectors) and preprocessing (scaling and mean-centering)

• supervised learning (support vector machine, model and feature selection, k-fold cross validation, overfitting)

• unsupervised learning (clustering, PCA, ICA)

• (I) size, weight and shape of biopolymers

• (II) atomic level structure of biopolymers (X-Ray crystallography)

• (III) atomic level structure of biopolymers (NMR spectroscopy)

• (IV) optical techniques

• Assessment of the stereochemical quality of protein structures

• Lack of experimental data and structure modeling from the sequence: homology modeling

• Protein-protein and protein-drug interactions (I): blind docking

• Protein-protein and protein-drug interactions (II): data driven docking

• Protein-protein association and dissociation by diffusion: brownian dynamics

• Molecules in motion (I): molecular dynamics of molecules and complexes in solution

• Molecules in motion (II): un/binding protein-ligand kinetics and thermodynamics

Lists, arrays, simple operations, examples, implementation

Stacks and queues (definition, examples, implementation)

Divide and conquer. Recursion. Greedy techniques (knapsack problem, radix sorting problem, Huffman codes). Binary search. Sorting algorithms. Sequence pattern analysis and matching. Algorithm optimization. Complexity analysis of algorithms

• Smith–Waterman

• Basic Local Alignment SearchTool (BLAST)

• Burrows-Wheeler Aligner (BWA)

• Bowtie and Burrows–Wheeler transform

• Bowtie2 and FM-index

• Quasi alignments (Salmon)

• Pseudo alignment (Kallisto)

BAM and CRAM

GTF, BED, BIGWIG, VCF

concepts, algorithms and implementation

paradigms; iterative, recursion, procedural, objects, functional. R and Python; Other such as Perl, Javscript, Ruby, ... where/when needed.

• Introduction

• the SQL query language

• Advance SQL

• Database Design: Entity Relationship Diagrams, Transactions, Indexes, Relational Algebra, Join

• when to use NoSQL and how they work

• Key-Value Store concepts

• Key-Value db, Redis a data structures server

• Document Store, Elastic

• advantages over RDBMS and NoSQL

• Graph db, Neo4j

• Graph db, Semantic Databases Triple stores

• Cloud: AWS DynamoDB Document and Key-Value

• Cloud: AWS Aurora

• Cloud: AWS Neptune

parametric vs non-parametric methods, binomial distribution, overdispersion, negative binomial model, multiple comparisons the false discovery rate (Bonferroni and Benjamini-Hochberg)

main types of enrichment analysis (list-based, ranked-based, list-based with relationship), statistics to test for enrichment (Fisher exact, hypergeometric, Kolmogorov-Smirnov, permutation)

DAVID, GSEA, GOstat, topGO, BINGO

hierarchical clustering, distance metrics, k-means