Advanced Genome Bioinformatics
UPF Master in Bioinformatics 2016-2017
eduardo.eyras [at] upf.edu
Probability and Classification
- AGB1: Description of the course (Slides)
- AGB2: Probabilities (Slides)
(Maximum Likelihood Exercise)
(Bayes' Theorem Exercise)
- AGB3: The Classification Problem. Bayesian Classification
- AGB4: Naive Bayes classifier. (Slides)
- AGB5: Evaluating of models.
- AGB6: Accuracy measures, ROC curves
- AGB7: Information, Entropy. Entropy-based measures.
- AGB8: (Slides)
- AGB9: Feature selection. Information Gain. Decision trees
- AGB11: Description of known Motifs. Modelling dependencies. Markov models.
- AGB12: (Slides)
Hidden Markov Models
- AGB13: HMMs: Definition. Examples.
- AGB14: The Viterbi, Forward and Posterior decoding algorithms. Viterbi Exercise.
- AGB15: Descripion of assignments
- AGB17: Work in class
- AGB19: Work in class
- AGB21: Work in class
- AGB23: Work in class
Assignment presentations: 15-20 mins presentations (all group participants should present).
Presentations should include a descriptions of the problem, methodologies used, results obtained and discussion/conclusions of the work.
- AGB25, AGB26:
- AGB27, AGB28:
- AGB29, AGB30:
- Naive Bayes model to classify tumor types gene expression patterns
- Markov model to predict RNA binding sites of a protein
- Hidden Markov model of 5' splice-site selection
- Hidden Markov model of the U2 branch point
- Hidden Markov model of the U12 branch point
Some Perl notes.
Other courses on-line
For Markov models, Hidden Markov Models and the EM algorithm, see the excellent courses by:
Some on-line PERL books and courses:
Some recommended reading on molecular biology and genetics: