You are on page 1of 11

Algorithms in

Bioinformatics
Lecture: 04
Instructor: Dr. M. Asif Rasheed
Algorithms

 Something that can be programmed and executed


 Computer
 Complexity classes of algorithms
 P: Set of problems solved by deterministic machine in polynomial time
 NP: Set of problems solved by non-deterministic machine in polynomial time
Algorithm design strategies

 Exhaustive search or brute force algorithm


 Searches through whole problem space to find the best solution
 Enumerating all possible candidates for the solution
 Whether each candidate satisfies the problem's statement

 Branch-and-Bound or pruning algorithm


 Explores branches of tree
 Subsets of the solution set
 Branch is checked
 Upper and lower bounds on the optimal solution
 Discarded if better solution is not found than the best one found so far by the
algorithm
 Dynamic Programming
 Problem is divided to small subproblems
 Solved and joined to solve the main problem
 Greedy algorithms
 Heuristic of making the locally optimal choice at each stage
 Choose most attractive path towards the solution at each iteration
 Traveling salesman problem
 At each stage visit an unvisited city nearest to the current city
 Not find a best solution, but terminates in a reasonable number of steps
 The Bridge Obsession Problem
 Find a tour crossing every bridge just once
 Eulerian Cycle Problem
 Given the graph in the image, is it possible to construct a path (or a cycle, i.e. a
path starting and ending on the same vertex) which visits each edge exactly once
 Find a cycle that visits every edge exactly once
 Linear time
 Hamiltonian Cycle Problem
 Find a cycle that visits every vertex exactly once
 NP– complete
Hidden Markov Models

 Simpler Markov models


 State is directly visible to observer
 The state transition probabilities are the only parameters
 Hidden Markov model
 State is not directly visible
 The output, dependent on the state, is visible
 Each state has a probability distribution
 Possible output tokens
HMM…

 Gene finding
 Finds gene through the specific labels
 HMM is trained to find them
 VEIL, HMMGene, Genie, etc
 Pattern recognition applications
 Protein coding regions in genome sequences
 Prediction of secondary structure elements from protein primary sequences
Smith-Waterman Algorithm

 The Smith–Waterman algorithm performs local sequence alignment


 For determining similar regions between two strings of sequences
 Compares segments of all possible lengths and optimizes the similarity
measure
 A dynamic programming algorithm
 It is guaranteed to find the optimal local alignment with scoring system
 Negative scoring matrix cells are set to zero
 Trace back procedure starts at the highest scoring matrix cell and proceeds
until a cell with score zero
Needleman–Wunsch algorithm

 One of the first applications of dynamic programming


 To compare biological sequences
 The algorithm was developed by Saul B. Needleman and Christian D. Wunsch
and published in 1970.
 The algorithm essentially divides a large problem into a series of smaller
problems
 Uses the solutions to the smaller problems to reconstruct a solution to the
larger problem.
 Widely used for optimal global alignment
Deadlock
11

You might also like