12
Bioinformatics • Protein structure prediction • Motif finding • Clustering techniques in bioinformatics • Sequence alignment and comparison • Phylogeny • Applying AI techniques such as Genetic Algorithm, Simulated Annealing, Neural Network, Support Vector Machine, etc

Bioinformatics

Embed Size (px)

DESCRIPTION

Bioinformatics. Protein structure prediction Motif finding Clustering techniques in bioinformatics Sequence alignment and comparison Phylogeny Applying AI techniques such as Genetic Algorithm, Simulated Annealing, Neural Network, Support Vector Machine, etc. - PowerPoint PPT Presentation

Citation preview

Page 1: Bioinformatics

Bioinformatics

• Protein structure prediction• Motif finding• Clustering techniques in bioinformatics• Sequence alignment and comparison• Phylogeny• Applying AI techniques such as Genetic

Algorithm, Simulated Annealing, Neural Network, Support Vector Machine, etc

Page 2: Bioinformatics

Simulation of Optical Multistage Networks

• Different Routing Algorithms

• Different Topologies

• Different Traffic (Poisson, self-similar)

• Graphical User Interface (Java)

• Animation (Java)

• Distributed Simulation (RMI in Java)

Page 3: Bioinformatics

• http://www.cs.gsu.edu/~cscyip/hong/omin.htm

• http://www.cs.gsu.edu/~matcyjx/thesis.html

Page 4: Bioinformatics

Non-blocking Networks

• Rearrangeable Non-blocking Networks

example: Benes network

• Can realize any permutation, but all the connection requests have to be known in advance (off-line).

• If not known in advance, existing paths have to rearranged if a conflict arises.

Page 5: Bioinformatics

Strictly Nonblocking

• Never have any conflicts even if connection requests are not known in advance.

• Crossbar is an example.

• Suitable for optical networks.

• Hardware is costly (O(N^2) hardware).

• Can we find a better network with hardware cost of O(N log N)?

Page 6: Bioinformatics

Questions?

• What kind of subsets are strictly nonblocking?

• What is the minimum number of rearrangements if conflicts arise in a rearrangeable non-blocking networks?

• Fault-tolerance

Page 7: Bioinformatics

Possible Projects:

• Simulation using general language (Java or JSIM)

• Simulate networks using OPNET package• Design and analysis of routing algorithms• Design new networks for better

performance (fewer crosstalks, short distance, fewer conversions, rearrange quickly, etc)

Page 8: Bioinformatics

Possible Projects

• Use neural networks to route messages

• Use genetic algorithms to schedule messages

• Use simulated annealing to route messages

• Use rule-based systems to do traffic control

• Management software

Page 9: Bioinformatics

Database Management for PCS Networks

• Personal Communication Service (PCS) Networks

• Mobile Communication

• Location Management –Home Location Register (HLR) and Visitor Location Register (VLR)

• How to track a Mobile Terminal (MT)?

Page 10: Bioinformatics

Mobile Networks

• Procedure for location registration

• New dynamic database management method

• Performance – Analytical and Simulation

• Simulation Program

• Efficient routing algorithms

• Find domination set (graph problem)

Page 11: Bioinformatics

Mobile Networks

• Channel Assignment Problem

• Frequency Reuse

• Fixed Channel Assignment

• Dynamic Channel Assignment

• Greedy Algorithms

• Genetic Algorithms

Page 12: Bioinformatics

Parallel Computing

• Solve engineering problems using MPI, PVM or OpenMP.

• Implement some CS problems (simulation, database, sequence alignment algorithms, data mining or neural networks) using distributed computing.

• Modeling and analysis of parallel task scheduling policies