1. www.bioalgorithms.infoAn Introduction to Bioinformatics Algorithms Dynamic Programming: Edit Distance

2. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info The Change Problem Goal: Convert some amount of money M into given denominations, using the fewest possible number of coins Input: An amount of money M, and an array of d denominations c = (c 1, c 2, …, c d ), in a decreasing order of value (c 1 > c 2 > … > c d ) Output: A list of d integers i 1, i 2, …, i d such that c 1 i 1 + c 2 i 2 + … + c d i d = M and i 1 + i 2 + … + i d is minimal

3. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: Example Given the denominations 1, 3, and 5, what is the minimum number of coins needed to make change for a given value? 12345678910 111 Value Min # of coins Only one coin is needed to make change for the values 1, 3, and 5

4. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: Example (cont ’ d) Given the denominations 1, 3, and 5, what is the minimum number of coins needed to make change for a given value? 12345678910 12121222 Value Min # of coins However, two coins are needed to make change for the values 2, 4, 6, 8, and 10.

5. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: Example (cont ’ d) 12345678910 1212123232 Value Min # of coins Lastly, three coins are needed to make change for the values 7 and 9 Given the denominations 1, 3, and 5, what is the minimum number of coins needed to make change for a given value?

6. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: Recurrence This example is expressed by the following recurrence relation: minNumCoins(M) = minNumCoins(M-1) + 1 minNumCoins(M-3) + 1 minNumCoins(M-5) + 1 min of

7. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: Recurrence (cont ’ d) Given the denominations c: c 1, c 2, …, c d, the recurrence relation is: minNumCoins(M) = minNumCoins(M-c 1 ) + 1 minNumCoins(M-c 2 ) + 1 … minNumCoins(M-c d ) + 1 min of

8. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Change Problem: A Recursive Algorithm 1.RecursiveChange(M,c,d) 2. if M = 0 3. return 0 4. bestNumCoins  infinity 5. for i  1 to d 6. if M ≥ c i 7. numCoins  RecursiveChange(M – c i, c, d) 8. if numCoins + 1 < bestNumCoins 9. bestNumCoins  numCoins + 1 10.return bestNumCoins

9. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info RecursiveChange Is Not Efficient It recalculates the optimal coin combination for a given amount of money repeatedly i.e., M = 77, c = (1,3,7): Optimal coin combo for 70 cents is computed 9 times!

10. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info The RecursiveChange Tree 74 77 76 70 757369737167696763 747268 72 70 66 686662 72 70 66 70 6864 666460 686662 666460 626056... 70

11. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info We Can Do Better We’re re-computing values in our algorithm more than once Save results of each computation for 0 to M This way, we can do a reference call to find an already computed value, instead of re-computing each time Running time M*d, where M is the value of money and d is the number of denominations

12. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info The Change Problem: Dynamic Programming 1.DPChange(M,c,d) 2. bestNumCoins 0  0 3. for m  1 to M 4. bestNumCoins m  infinity 5. for i  1 to d 6. if m ≥ c i 7. if bestNumCoins m – c i + 1 < bestNumCoins m 8. bestNumCoins m  bestNumCoins m – c i + 1 9. return bestNumCoins M

13. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info DPChange: Example 0 01 012 0123 01234 012345 0123456 01234567 012345678 0123456789 01 0 012 0121 01212 012123 0121232 01212321 012123212 0121232123 c = (1,3,7) M = 9

14. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Manhattan Tourist Problem (MTP) Imagine seeking a path (from source to sink) to travel (only eastward and southward) with the most number of attractions (*) in the Manhattan grid Sink * * * * * * * ** * * Source *

15. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Manhattan Tourist Problem (MTP) Imagine seeking a path (from source to sink) to travel (only eastward and southward) with the most number of attractions (*) in the Manhattan grid Sink * * * * * * * ** * * Source *

16. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Manhattan Tourist Problem: Formulation Goal: Find the longest path in a weighted grid. Input: A weighted grid G with two distinct vertices, one labeled “source” and the other labeled “sink” Output: A longest path in G from “source” to “sink”

17. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: An Example 324 073 33 0 132 4 4 5 6 4 6 5 5 8 2 2 5 0123 0 1 2 3 j coordinate i coordinate 13 source sink 4 3240 10 2 43 3 1 1 2 2 2 4 19 95 15 23 0 20 3 4

18. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Greedy Algorithm Is Not Optimal 125 2 1 5 23 4 000 5 3 0 3 5 0 10 3 5 5 1 2 promising start, but leads to bad choices! source sink 18 22

19. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Simple Recursive Program MT(n,m) if n=0 or m=0 return MT(n,m) x  MT(n-1,m)+ length of the edge from (n- 1,m) to (n,m) y  MT(n,m-1)+ length of the edge from (n,m-1) to (n,m) return max{x,y}

20. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Simple Recursive Program MT(n,m) x  MT(n-1,m)+ length of the edge from (n- 1,m) to (n,m) y  MT(n,m-1)+ length of the edge from (n,m-1) to (n,m) return min{x,y} What’s wrong with this approach?

21. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info 1 5 01 0 1 i source 1 5 S 1,0 = 5 S 0,1 = 1 Calculate optimal path score for each vertex in the graph Each vertex’s score is the maximum of the prior vertices score plus the weight of the respective edge in between MTP: Dynamic Programming j

22. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Dynamic Programming (cont ’ d) 12 5 3 012 0 1 2 source 13 5 8 4 S 2,0 = 8 i S 1,1 = 4 S 0,2 = 3 3 -5 j

23. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Dynamic Programming (cont ’ d) 12 5 3 0123 0 1 2 3 i source 13 5 8 8 4 0 5 8 103 5 -5 9 13 1-5 S 3,0 = 8 S 2,1 = 9 S 1,2 = 13 S 3,0 = 8 j

24. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Dynamic Programming (cont ’ d) greedy alg. fails! 125 -51 3 0 5 3 0 3 5 0 10 -3 -5 0123 0 1 2 3 i source 138 5 8 8 4 9 138 9 12 S 3,1 = 9 S 2,2 = 12 S 1,3 = 8 j

25. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Dynamic Programming (cont ’ d) 125 -51 33 00 5 3 0 3 5 0 10 -3 -5 2 0123 0 1 2 3 i source 138 5 8 8 4 9 138 12 9 15 9 j S 3,2 = 9 S 2,3 = 15

26. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info MTP: Dynamic Programming (cont ’ d) 125 -51 33 00 5 3 0 3 5 0 10 -3 -5 2 0123 0 1 2 3 i source 138 5 8 8 4 9 138 12 9 15 9 j 0 1 16 S 3,3 = 16 (showing all back-traces) Done!

27. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment: 2 row representation Alignment : 2 * k matrix ( k > m, n ) AT--GTAT ATCG A C letters of v letters of w T T ATCTGAT TGCATA v : w : m = 7 n = 6 4 matches2 insertions2 deletions Given 2 DNA sequences v and w:

28. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Aligning DNA Sequences V = ATCTGATG W = TGCATAC n = 8 m = 7 ATCTGATG TGCATAC V W match deletion insertion mismatch indels 4 1 2 2 matches mismatches insertions deletions

29. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Longest Common Subsequence (LCS) – Alignment without Mismatches Given two sequences v = v 1 v 2 …v m and w = w 1 w 2 …w n The LCS of v and w is a sequence of positions in v : 1 < i 1 < i 2 < … < i t < m and a sequence of positions in w : 1 < j 1 < j 2 < … < j t < n such that i t -th letter of v equals to j t -letter of w and t is maximal

30. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info LCS: Example AT -- CTGATC TGCT A C elements of v elements of w -- A 1 2 0 1 2 2 3 3 4 3 5 4 5 5 6 6 6 7 7 8 j coords: i coords: Matches shown in red positions in v: positions in w: 2 < 3 < 4 < 6 < 8 1 < 3 < 5 < 6 < 7 Every common subsequence is a path in 2-D grid 0 0 (0,0)  (1,0)  (2,1)  (2,2)  (3,3)  (3,4)  (4,5)  (5,5)  (6,6)  (7,6)  (8,7)

31. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info LCS: Dynamic Programming Find the LCS of two strings Input: A weighted graph G with two distinct vertices, one labeled “source” one labeled “sink” Output: A longest path in G from “source” to “sink” Solve using an LCS edit graph with diagonals replaced with +1 edges

32. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info LCS Problem as Manhattan Tourist Problem T G C A T A C 1 2 3 4 5 6 7 0i ATCTGATC 012345678 j

33. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Graph for LCS Problem T G C A T A C 1 2 3 4 5 6 7 0i ATCTGATC 012345678 j

34. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Graph for LCS Problem T G C A T A C 1 2 3 4 5 6 7 0i ATCTGATC 012345678 j Every path is a common subsequence. Every diagonal edge adds an extra element to common subsequence LCS Problem: Find a path with maximum number of diagonal edges

35. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Computing LCS Let v i = prefix of v of length i: v 1 … v i and w j = prefix of w of length j: w 1 … w j The length of LCS(v i,w j ) is computed by: s i, j = max s i-1, j s i, j-1 s i-1, j-1 + 1 if v i = w j

36. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Computing LCS (cont ’ d) s i,j = MAX s i-1,j + 0 s i,j -1 + 0 s i-1,j -1 + 1, if v i = w j i,j i -1,j i,j -1 i -1,j -1 1 0 0

37. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Every Path in the Grid Corresponds to an Alignment 01234 0 1 2 3 4 W ATCG A T G T V 0 1 2 2 3 4 V = A T - G T | | | W= A T C G – 0 1 2 3 4 4

38. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Aligning Sequences without Insertions and Deletions: Hamming Distance Given two DNA sequences v and w : v : The Hamming distance: d H (v, w) = 8 is large but the sequences are very similar ATATATAT ATATATATw :w :

39. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Aligning Sequences with Insertions and Deletions v :ATATATAT ATATATATw :w :-- By shifting one sequence over one position: The edit distance: d H (v, w) = 2. Hamming distance neglects insertions and deletions in DNA

40. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance Levenshtein (1966) introduced edit distance between two strings as the minimum number of elementary operations (insertions, deletions, and substitutions) to transform one string into the other d(v,w) = MIN number of elementary operations to transform v  w

41. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance vs Hamming Distance V = ATATATAT W = TATATATA Hamming distance always compares i -th letter of v with i -th letter of w Hamming distance: d(v, w)=8 Computing Hamming distance is a trivial task.

42. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance vs Hamming Distance V = ATATATAT W = TATATATA Hamming distance: Edit distance: d(v, w)=8 d(v, w)=2 Computing Hamming distance Computing edit distance is a trivial task is a non-trivial task W = TATATATA Just one shift Make it all line up V = - ATATATAT Hamming distance always compares i -th letter of v with i -th letter of w Edit distance may compare i -th letter of v with j -th letter of w

43. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance vs Hamming Distance V = ATATATAT W = TATATATA Hamming distance: Edit distance: d(v, w)=8 d(v, w)=2 (one insertion and one deletion) How to find what j goes with what i ??? W = TATATATA V = - ATATATAT Hamming distance always compares i -th letter of v with i -th letter of w Edit distance may compare i -th letter of v with j -th letter of w

44. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance: Example TGCATAT  ATCCGAT in 5 steps TGCATAT  (delete last T) TGCATA  (delete last A) TGCAT  (insert A at front) ATGCAT  (substitute C for 3 rd G) ATCCAT  (insert G before last A) ATCCGAT (Done)

45. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance: Example TGCATAT  ATCCGAT in 5 steps TGCATAT  (delete last T) TGCATA  (delete last A) TGCAT  (insert A at front) ATGCAT  (substitute C for 3 rd G) ATCCAT  (insert G before last A) ATCCGAT (Done) What is the edit distance? 5?

46. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance: Example (cont ’ d) TGCATAT  ATCCGAT in 4 steps TGCATAT  (insert A at front) ATGCATAT  (delete 6 th T) ATGCATA  (substitute G for 5 th A) ATGCGTA  (substitute C for 3 rd G) ATCCGAT (Done)

47. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Edit Distance: Example (cont ’ d) TGCATAT  ATCCGAT in 4 steps TGCATAT  (insert A at front) ATGCATAT  (delete 6 th T) ATGCATA  (substitute G for 5 th A) ATGCGTA  (substitute C for 3 rd G) ATCCGAT (Done) Can it be done in 3 steps???

48. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info The Alignment Grid Every alignment path is from source to sink

49. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment as a Path in the Edit Graph 0 1 2 2 3 4 5 6 7 7 A T _ G T T A T _ A T _ G T T A T _ A T C G T _ A _ C A T C G T _ A _ C 0 1 2 3 4 5 5 6 6 7 (0,0), (1,1), (2,2), (2,3), (3,4), (4,5), (5,5), (6,6), (7,6), (7,7) - Corresponding path -

50. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignments in Edit Graph (cont’d) and represent indels in v and w with score 0. represent matches with score 1. The score of the alignment path is 5.

51. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment as a Path in the Edit Graph Every path in the edit graph corresponds to an alignment:

52. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment as a Path in the Edit Graph Old Alignment 0122345677 0122345677 v= AT_GTTAT_ w= ATCGT_A_C 0123455667 0123455667 New Alignment 0122345677 0122345677 v= AT_GTTAT_ w= ATCG_TA_C 0123445667 0123445667

53. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment as a Path in the Edit Graph 0122345677 0122345677 v= AT_GTTAT_ w= ATCGT_A_C 0123455667 0123455667 (0,0), (1,1), (2,2), (2,3), (3,4), (4,5), (5,5), (6,6), (7,6), (7,7)

54. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment: Dynamic Programming s i,j = s i-1, j-1 +1 if v i = w j max s i-1, j s i, j-1

55. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Dynamic Programming Example Initialize 1 st row and 1 st column to be all zeroes. Or, to be more precise, initialize 0 th row and 0 th column to be all zeroes.

56. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Dynamic Programming Example S i,j = S i-1, j-1 max S i-1, j S i, j-1  value from NW +1, if v i = w j  value from North (top)  value from West (left)

57. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment: Backtracking Arrows show where the score originated from. if from the top if from the left if v i = w j

58. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Backtracking Example Find a match in row and column 2. i=2, j=2,5 is a match (T). j=2, i=4,5,7 is a match (T). Since v i = w j, s i,j = s i-1,j-1 +1 s 2,2 = [s 1,1 = 1] + 1 s 2,5 = [s 1,4 = 1] + 1 s 4,2 = [s 3,1 = 1] + 1 s 5,2 = [s 4,1 = 1] + 1 s 7,2 = [s 6,1 = 1] + 1

59. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Backtracking Example Continuing with the dynamic programming algorithm gives this result.

60. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment: Dynamic Programming s i,j = s i-1, j-1 +1 if v i = w j max s i-1, j s i, j-1

61. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Alignment: Dynamic Programming s i,j = s i-1, j-1 +1 if v i = w j max s i-1, j +0 s i, j-1 +0 This recurrence corresponds to the Manhattan Tourist problem (three incoming edges into a vertex) with all horizontal and vertical edges weighted by zero.

62. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info LCS Algorithm 1.LCS(v,w) 2. for i  1 to n 3. s i,0  0 4. for j  1 to m 5. s 0,j  0 6. for i  1 to n 7. for j  1 to m 8. s i-1,j 9. s i,j  max s i,j-1 10. s i-1,j-1 + 1, if v i = w j 11. “ “ if s i,j = s i-1,j b i,j  “ “ if s i,j = s i,j-1 “ “ if s i,j = s i-1,j-1 + 1 return (s n,m, b)

63. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Now What? LCS(v,w) created the alignment grid Now we need a way to read the best alignment of v and w Follow the arrows backwards from sink

64. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info Printing LCS: Backtracking 1.PrintLCS(b,v,i,j) 2. if i = 0 or j = 0 3. return 4. if b i,j = “ “ 5. PrintLCS(b,v,i-1,j-1) 6. print v i 7. else 8. if b i,j = “ “ 9. PrintLCS(b,v,i-1,j) 10. else 11. PrintLCS(b,v,i,j-1)

65. An Introduction to Bioinformatics Algorithmswww.bioalgorithms.info LCS Runtime It takes O(nm) time to fill in the nxm dynamic programming matrix. Why O(nm)? The pseudocode consists of a nested “for” loop inside of another “for” loop to set up a nxm matrix.