Scope of the course:

This course investigates the following four interconnected components: formal grammar systems, stochastic models, optimization algorithms, and biomolecular patters and structures. The study will be focused on stochastic grammars to model biomolecules and algorithms to compute with such models for the purpose of pattern and structure prediction.

More description of the course.

1. Papers 4, 5, 6 and 7 following this link [Eddy]
   
2. Pioneer work in SCFG for RNA secondary structure prediction [UCSC group]
   
3. Lightweight SCFGs for RNA secondary structure prediction [Eddy group]
   
4. The language of RNA: a formal grammar that includes pseudoknots [Eddy and Rivas]
   
5. Complex probabilistic models for RNA secondary structure prediction [Eddy group]
   
6. Subclasses of tree-adjoining grammars for RNA secondary structure [Uemura group]
   
7. Tree adjoining grammars for RNA pseudoknots [Uemura group]
   
8. Parallel grammar systems for RNA pseudoknots [Cai and Malmberg group]
   
9. Computational linguistics of biological sequences [Searls group]
   
10. Grammatical representations of Macromolecular structure [Chiang and Searls]
    
11. Computational linguistics for biopolymer structures and statistical mechanics [Dill et al]
    
12. Routes are trees: parsing perspectives on protein folding [Dill group]
    
13. Graph grammar modeling RNA tertiary structure motifs [Major group]
    
14. Stochastic k-tree grammar for biomolecular structure modeling [Cai and Malmberg group]
    
15. PPT in non-coding RNA structure, modeling, prediction, and search [Cai's presentation]
    
16. A ppt presentation in SCFG [from the web]
    

Projects

• Project 1

1. Chomsky grammars
   regular, linear and context-free grammars, derivation, ambiguity

2. Stochastic grammar models
   HMM, SCFG, computation with models

3. Parsing algorithms
   decoding, probability computation, model learning

4. Biomolecular patters and structures
   DNA motif detection, RNA secondary structure prediction

5. Grammar systems for context-sensitivity
   non-Chomsky grammar, mildly CSG

6. Modeling of bimolecular higher order structures
   applications in protein and RNA tertiary structures

The time allocated for each section will be roughly equally, 14/6 weeks. The teaching will be a mix of lectures by the instructor and presentations by students on their literature-readings and research projects. No textbook will be used.

Grading will be based on project reports (40%), presentations (40%), and participation in classroom discussions (20%).

CSCI 2670 (Theory of Computation), or CSCI 4470/6470 (Algorithms), or CSCI 4490/6490 (Algorithms for Computational Biology), or the approval of the instructor. No prior knowledge, but an interest, in molecular biology is essential for this course.

It is expected that the work you submit is your own. Plagiarism and other forms of academic dishonesty will be handled within the guidelines of the Student Handbook. The usual penalty for academic dishonesty is loss of credit for the assignment in question; however, stronger measures may be taken when conditions warrant.