| Date | Lecture Topic(s) | Due |
|---|---|---|
| Fri 23 Sept | administrivia, fundamental dogma, DNA | |
| Mon 26 Sept | models: DNA AT pair, protein backbone, alpha helices, sidechains GAILV | |
| Wed 28 Sept | models: beta sheets parallel/antiparallel, type I/II hairpin turns, residues ST | |
| Fri 30 Sept | multiple alignments, A2M format, models: CHDEFYW, disulfides, coordinating metal ions, aromatics | |
| Mon 3 Oct | Phoenix Eagleshadow (fellowships); codons, stochastic models as computable probability functions, Bayes rule | |
| Wed 5 Oct | zero-order and first-order Markov chains; stop characters; factoring out length model; mention of higher order chains | |
| Fri 7 Oct | log-probability for computation, efficient addition of probs in log-prob space, train/crosstrain/test, n-fold cross-validation | perl 1 (noon) |
| Mon 10 Oct | setting parameters for zero-order Markov model, derivation of maximum-likelihood estimate (MLE) using Lagrange multipliers, pseudocounts as mean a posteriori (MAP) estimate without derivation, use of zero-order model as pseudocounts for first-oder model, notion of P-value (for single trial) and E-value | |
| Wed 12 Oct | review of returned homework; questions on Darling models (cis and trans conformations, CDA and DAC angles); P1, Pn, and E-value; equivalence of E and Pn for small E | |
| Fri 14 Oct | Z-scores; Gaussian vs. Extreme-value (Gumbel) distribution; pdf and cdf for extreme value; derivation of linear tail; Shannon entropy; information gain | Darling models (noon) |
| Mon 17 Oct | information gain, entropy and relative entropy, Kullbak-Leibler divergence, (contrasted relative entropy and difference in entropy), sequence logos | |
| Wed 19 Oct | sequence logos, mutual information, ROC plots | |
| Fri 21 Oct | ROC plots, sensitivity, selectivity, precitions, ROCn plots, started discussion on sequence alignment | Fellowship application (noon) |
| Mon 24 Oct | Subsitution matrices; scoring gapless alignments; diferent gap cost models | |
| Wed 26 Oct | global alignment arbitrary gap costs and traceback; global alignment with linear gap costs | |
| Fri 28 Oct | feedback on fellowship applications; global alignment with linear and affine gap costs; traceback on linear gap costs | perl 2 (noon) |
| Mon 31 Oct | global and local alignment for affince gap costs, with traceback | |
| Wed 2 Nov | Intro to hidden Markov models | |
| Fri 4 Nov | Catherine Soehner and Christy Hightower (Science Library) | perl 3 (noon) |
| Mon 7 Nov | David Kulp (U. Mass. Amherst)
Causal inference of regulator-target pairs by gene mapping of expression phenotypes
The combination of whole genome expression profiling and polymorphic marker screening has emerged as an ideal genetic perturbation model to detect causal relationships among genes. By treating expression as a quantitative phenotype, linkage analysis can reveal associated regulatory loci. We developed an epistatic-like linkage model to jointly account for gene expression and genotype and precisely map regulator genes. A consideration of complete and reduced forms of the model provides the means to dissect regulator-target relationships as causal or merely dependent. In simulations we find that the model is robust with respect to multiple independent regulators and we show that, in yeast, regulator genes are accurately predicted and that regulatory modules derived from pairwise linkage have biological significance. | |
| Wed 9 Nov | review of HMM forward algorithm; HMM backward algorithm to get P(state s in position i|sequence) | |
| Fri 11 Nov | NO CLASS (Veteran's Day) | |
| Mon 14 Nov | Quick review of forward-backward. Baum-Welch training. Profile HMMs (both HMMer's Plan 7 and SAM 9-edges) | perl 4 (noon) |
| Wed 16 Nov | Intro to Dirichlet mixtures | |
| Fri 18 Nov | More on Dirichlet mixtures | perl 5 (noon) |
| Mon 21 Nov | Multiple alignment. The problem, the progressive alignment heuristic. Problems with current benchmarks (favoring overalignment) | |
| Wed 23 Nov | Guest Lecture: Melissa Cline
Visualizing the effects of splice variation on protein interaction networks High-throughput technologies have generated large networks of protein-protein interactions. These networks are a powerful aid to expression analysis, as they relate the expression values of individual genes in a framework that describes functional association, and perhaps causation. Cytoscape is an open-source package for visualization and modeling of protein interaction networks, with support for visualizing expression results from one or more experiments. Yet splice variation complicates this picture by removing interaction domains, and thus altering the network topology. Most network analysis packages do not take this into account. This talk describes work underway at the Pasteur and Max-Planck Institutes for translating protein interaction networks into protein domain interaction networks, and integrating these networks with exon expression results from the Affymetrix Human Exon array. | |
| Fri 25 Nov | NO CLASS | |
| Mon 28 Nov | Evolutionary trees. Paralogs vs. orthologs. %ID, % similar, converting identity to distance. Additivity assumption. UPGMA algorithm | web and literature search (noon) |
| Wed 30 Nov | Evolutionary trees. Ultrametric distances, UPGMA, and neighbor-joining. Brief mentions of maximum likelihood and parsimony. | |
| Fri 2 Dec | Stochastic context-free grammars and RNA structure | perl 6 noon |
| Wed 7 Dec 4--7 p.m. | exam period (NOT USED) |
|
|
| BME 205 home page | Karplus's lab page | UCSC Bioinformatics research |
Questions about page content should be directed to
Kevin Karplus
Biomolecular Engineering
University of California, Santa Cruz
Santa Cruz, CA 95064
USA
karplus@soe.ucsc.edu
1-831-459-4250