Alignment of Pairs of Sequences 1

1
Alignment of Pairs of Sequences (1)

• BCI 3043 Introduction to Bioinformatics

2

• Biosequence alignment problems
• Global alignment
• Local alignment
• Gap penalty
• Affine gap penalty
• Methods of sequence alignment
• Dot matrix
• Dynamic programming
• Heuristic
• Scoring matrices
• PAM
• BLOSUM
• GOMET
• Efficiency measurement

3
Genes encode the recipes for proteins
4
Proteins Molecular Machines

• Proteins in your muscles allows you to
  movemyosinandactin

5
Intro

• Pairwise sequence alignment process of
  comparing 2 sequences by searching for a series
  of individual character that are in the same
  order in the sequence
• Compare 2 sequence (DNA_at_Protein) and find the
  similarities
• DNA alignment can give ambiguous result
• Translate into protein before aligning them!!

6
Intro

• Consider two sequence X and Y
• AGGCTATCACCTGACCTCCAGGCCGATGCCC sequence X
• TAGCTATCACGACCGCGGTCGATTTGCCCGAC sequence Y
• We assign gap to position 0.n in X and Y
• -AGGCTATCACCTGACCTCCAGGCCGA--TGCCC---
• TAG-CTATCAC--GACCGC--GGTCGATTTGCCCGAC
• And we get two sequence with same length and as
  similar to each other as possible
• But, how much the similarity is?
• Give score to matches, mismatches and gaps

7
Why to do alignment?

• Hypothesis all genetic material had one
  ancestral
• But, evolution and mutation has modified and
  create differences
• Insertion, deletion, substitution
• Indel insertion and deletion

8
Cont.

• Comparison of macromolecular sequences.
• Nucleic acids (DNA, RNA) or proteins.
• Suggest evolutionary, structural and functional
  relationships.
• Heuristic algorithms for practical database
  searching.

9
Cont.

• The draft human genome is available
• Automated gene finding is possible
• Gene AGTACGTATCGTATAGCGTAA
• What does it do?
• One approach Is there a similar gene in another
  species?
• Align sequences with known genes
• Find the gene with the best match

10
Distance vs Similarity

• Distance minimal sum of weights for a set of
  mutation transforming one to another
• Similarity ???
• Similarity useful for db searching and distance
  useful for phylogenetic tree construction
• Maximizing similarities minimizing distances

11
Homology

• Common evolutionary relationship between two
  individual
• Indicates by a high level of similarities
• 25 identity over 100 amino acids meant these two
  sequences has a common ancestry
• Homology is different with similarity

12
Biological motivations of alignment problems

• 5 major variants
• Global alignments
• Local alignments
• Ends free-space alignment
• Gap penalty
• Affine gap penalty

13
1) Global alignment

• To find similarities between sequence S and T
• Done across entire sequence length
• E.g Needleman-Wunsch algorithm
• EMBOSS http//www.hgmp.mrc.ac.uk/Software/EMBOSS/A
  ps/needle.html

14
Global Alignment

• Input two sequences over the same alphabet
• Output an alignment of the two sequences
• Example
• GCGCATGGATTGAGCGA and TGCGCCATTGATGACCA
• A possible alignment
• -GCGC-ATGGATTGAGCGA
• TGCGCCATTGAT-GACC-A

15
Global Alignment
Example (cont)

• -GCGC-ATGGATTGAGCGA
• TGCGCCATTGAT-GACC-A
• Three elements
• Perfect matches
• Mismatches
• Insertions deletions (indel)

Symmetric view of evolution
16
Global Alignmentscoring scheme

• Score each position independently
• Match 1
• Mismatch -1
• Indel -2
• Score of an alignment is sum of position scores

Example -GCGC-ATGGATTGAGCGA TGCGCCATTGAT-GAC
C-A Score (1x13) (-1x2) (-2x4)
3 ------GCGCATGGATTGAGCGA TGCGCC----ATTGATGACCA--
Score (1x5) (-1x6) (-2x11) -23
17
2) Local alignment

• Definition
• S1 , S2 strings. Find substrings a , ß of S1 ,
  S2 , whose similarity is maximum over all pairs
  of substrings from S1 , S2.
• Search the substring in two sequences that most
  similar
• More meaningful than global
• Useful in potential sequence and functional motif
  detection
• Cant get overall similarity
• E.g Smith-Waterman algorithm
• EMBOSS
• http//www.hgmp.mrc.ac.uk/Software/EMBOSS/Aps/wate
  r.html

18
Local Alignment

• Consider sequence A and B
• A aaaacccccggggtta
• B ttcccgggaaccaacc
• Only red region is similar
• If
• A aaaaccccgggtta
• B ttcccgggaaccaacc
• then
• ccc gg
• cccggg
• A aaaacccc ggctta
• B ttcccgggaaccaacc

19
3) Gap penalty

• Consider sequence P and Q with gap
• ACGTCTGATACGCCGTATAGTCTATCT Seq P
• ACGTCTGAT-------ATAGTCTATCT Seq Q
• We penalize gap using n x d where n number of
  gaps and d is gap penalty
• But, gap usually in bunch
• More realistic approach, reduce the gap penalty
  as the gap length increase

20
Affine gap penalty

• Example of gap penalty method
• Each gap is given weight depends on its length
• Wx g rx
• Wx total weight, g opening penalty, r
  gap penalty for each element, x gap length
• Example Genematcher2
• http//www.paracel.com/products/gm2.html

21
(No Transcript)
22
Group presentation

• Wednesday, 12/08/07
• gtlt10 minutes per group
• Smith-waterman overview
• Smith-waterman algorithm
• Needleman-Wunsch overview
• Needleman-Wunsch algorithm
• Affine gap penalty