Dynamic programming
Download as PPTX, PDF43 likes30,887 views
Dynamic programming is used for sequence alignment and other bioinformatics tasks. It works by breaking problems down into smaller subproblems. Needleman-Wunsch introduced an algorithm for global sequence alignment using dynamic programming that maximizes matches between sequences. The algorithm involves initializing a matrix, filling it using scoring schemes, and backtracking to trace alignments. Local alignment follows a similar approach but replaces negative values in the matrix with zeros to restrict alignments.
Dynamic programming
- 1. MARYAM BIBI FA12-BTY-011 TOPIC : DYNAMIC PROGRAMING SUBJECT : BIOINFIRMATICS
- 2. Alignment used to uncover homologies between sequences combined with phylogenetic studies can determine orthologous and paralogous relationships Global Alignments compares one whole sequence with other entire sequence computationally expensive Local Alignment uses a subset of a sequence and attempts to align it to subset of other sequences computationally less expensive
- 3. Dynamic Programming: dynamic programing is solving complex prblems by breaking them into a simpler subproblems. Problem can be divided into many smaller parts. Needleman and Wunsch were the first to propose this method.
- 4. Needleman and Wunsch describes general algorithm for sequence aignment. Maximize a score of similarity to give maximun match. Maximun match= largest number of nucleotides that can be match with others. That want to quantify sequence similarity between two sequences.
- 5. Dynamic programming in bioinformatics Dynamic programming is widely used in bioinformatics for the tasks such as sequence alignment, protein folding, RNA structure prediction and protein-DNA binding. First dynamic programming algorithms for protein-DNA binding were developed in the 1970s independently by Charles Delisi in USA and Georgii Gurskii and Alexanderr zasedatelev in USSR.
- 6. Dynamic Programming in sequence alignment There are three steps in dynamic programing. 1. initialization. The first step in the global alignment dynamic programming approach is to create a matrix with M + 1 columns and N + 1 rows where M and N correspond to the size of the sequences to be aligned. 2. Matrix filling(scoring) We fill the matrix with highest possible score. To align with diagnol ( align in next position.) Allign in off-diagonal requires inserion of crossponding gaps 3. Traceback and aligning Move from last corner and follow arrow.
- 7. Global alignment via dynamic programing 1st column and 1st row will be empty. Fill 1st block with zero Then fill 1st row and 1st coulmn with gap penality multiples. While filling the matrix there are three possible values horizental: score+ gap penality Verticle: score+ gap penality Diagonal: score+(match/mismatch) We have to write max score from these values in a cell Let , match = + 1 mismatch= -1 gap penality= -2
- 8. Lets Seq# 1 A A A C Seq# 2 A G C -
- 9. A A T C 0 -2 -4 -6 -8 A -2 ( 1) ( -4) 1 (-4) -1 -6 -1 -1 -3 -5 G -4 -1 0 -2 -4 C -6 -3 -2 -1 -1
- 10. Backward tracking In backward tracking we have to move from last cell( lower corner) and follows arrow from which cell the current cell’s values come from and go ahead.
- 11. A A T C 0 -2 -4 -6 -8 A -2 1 -1 -3 -5 G -4 -1 0 -2 -4 C -6 -3 -2 -1 -1
- 12. A A T C 0 -2 -4 -6 -8 A -2 1 -1 -3 -5 G -4 -1 0 -2 -4 C -6 -3 -2 -1 -1 Path 1 Path 2
- 13. Backtracking Now we have to allign this sequence. For alligning there are 2 rules. 1. If the value come from column we will have to write two sequences 2. If value come from horizontal or vertical then we will have to write perpendicular and add gap to other side.
- 14. Backtracking Here we have two paths. So we will get two possible sequences. We will write sequence from 3’-end. Path 1: Seq#1 A A T C Seq#2 - A G C
- 15. Backtracking Path 2 Seq# 1 A A T C Seq# 2 A G - C Hense , we do global allignment in this way.
- 16. Local alignment via dynamic programing Algorithim is same as in global alignment, but there are some changes. We fill 1st column and 1st row with zero. If the value comes in negative number than it is replaced by zero. Backtracking will be start from maximun value.
- 17. Local alignment Let, Match = 1 Mismatch = 0 Gap penality= 0 And sequence#1 GAATTCAGTTA Squence#2 GGATCGA
- 18. G A A T T C A G T T A 0 0 0 0 0 0 0 0 0 0 0 0 G 0 1 1 1 1 1 1 1 1 1 1 1 G 0 1 1 1 1 1 1 1 2 2 2 2 A 0 1 2 2 2 2 2 2 2 2 2 3 T 0 1 2 2 3 3 3 3 3 3 3 3 C 0 1 2 2 3 3 4 4 4 4 4 4 G 0 1 2 2 3 3 4 4 5 5 5 5 A 0 1 2 3 3 3 3 4 5 5 5 6
- 19. Bactracking After the matrix fill step, the maximum alignment score for the two test sequences is 6. The traceback step determines the actual alignment(s) that result in the maximum score.
- 20. G A A T T C A G T T A 0 0 0 0 0 0 0 0 0 0 0 0 G 0 1 1 1 1 1 1 1 1 1 1 1 G 0 1 1 1 1 1 1 1 2 2 2 2 A 0 1 2 2 2 2 2 2 2 2 2 3 T 0 1 2 2 3 3 3 3 3 3 3 3 C 0 1 2 2 3 3 4 4 4 4 4 4 G 0 1 2 2 3 3 4 4 5 5 5 5 A 0 1 2 3 3 3 3 4 5 5 5 6
- 21. Backtracking Rule will be same for this as in global alignment Sequence alignment : Seq#1 G A A T T C A G T TA Seq#2 G A - T C – G - - A So in this way we align the sequence using dynamic programing.