CISC667, F05, Lec27, Liao1 CISC 667 Intro to Bioinformatics (Fall 2005) Review Session

CISC667, F05, Lec27, Liao2 Basics of Molecular biology Central dogma –Transcription –Translation Genetic code DNA –Double helix –Watson-Crick binding RNA –Secondary structures Genes –Compositional Structure –Reading frames Proteins –Secondary structure: alpha helices, beta sheets, and coils –3-d structure Gene Regulations (Operons) DNA Microarray Cloning PCR Gel electrophoresis Yeast 2 hybrid system

CISC667, F05, Lec27, Liao3 Computational methods Dynamic programming –Problems can be dissect to sub-problems –Recurrence equations –DP Table –Traceback Maximum Likelihood Pairwise alignments Multiple alignments Hidden Markov models - viterbi, forward, and baum-welch (Maximum Likelihood) Phylogenetic trees - distance-based, character-based, and probability-based - Bootstrap Structural predictions - secondary - lattice model Clustering Profiles and inferring genetic networks K-means Support vector machines

CISC667, F05, Lec27, Liao4 Support Vector Machines Supervised learning Concepts and assumptions –Vectorization –Separating hyperplane with maximum margin –Feature space (high dimension) –Dot product, Kernel functions –Support vectors Algorithms –Rosenblatt’s for linearly separable data Prime form Dual form –Rosenblatt’s for non-linearly separable data (using kernels)

CISC667, F05, Lec27, Liao5 RNA secondary structures Concepts –Single strand –Intramolecular pairings –Structural categories Hairpins, bulge loops, interior loops, and multi loops Pseudo knots –Similarity in secondary structure plays a more important role than primary sequence in determining homology for RNAs Algorithms for predicting structures –Nussinov’s Maximum base pairing –Zuker’s Minimum energy

CISC667, F05, Lec27, Liao6 Protein structures Concepts –Primary, secondary, tertiary, and quaternary Algorithms for predicting structures –Secondary structure encoding Artificial neural networks Support vector machines –3-dimensional Lattice models –HP models: Monte Carlo, Genetic algorithm, and hybrid

CISC667, F05, Lec27, Liao7 Genetic networks Concepts –Regulation, protein-protein interactions, … –Operons Modeling –Boolean networks Acyclic Truth tables 3 steps predictor Maximum entropy chooser

CISC667, F05, Lec27, Liao8 About the Final Exam –Time and Place: December 9, 7:00PM-9:00PM 308 Gore Hall –closed-book –Four parts Primary sequence analysis [35 points] Support vector machines [25 points] Structural prediction [20 points] Clustering gene expression profiles (K-means) and inferring genetic network [20 points]