Lecture Slides on Two Sequence Alignment and Scoring Models | BME 110, Assignments of Chemistry

Download Lecture Slides on Two Sequence Alignment and Scoring Models | BME 110 and more Assignments Chemistry in PDF only on Docsity!

Two Sequence Alignment &^ Scoring Matrices

BME 110: CompBio Tools

Todd Lowe April 08, 2008

Admin

• Reading:^ – Chapter 3 should be completed^ – Chapter 5 for tuesday • Homework #1 due tomorrow (Fri) 5pm • Homework #2 assigned Tuesday

Full GenomeDot-Plot

Multiple Genome Alignment

Dot Plots P.calidifontis

P.arsenaticum

P.islandicum P.aerophilum^

P.aerophilum^

P.aerophilum

P.calidifontis

P.islandicum

P.islandicum P.arsenaticum^

P.arsenaticum^

P.calidifontis

Pair-wise Sequence Comparison • Basis for relating biological information from awell-studied gene to a new sequence • Many programs exist for pairwise comparison • Some specialize in fast

database searching and get “good” alignments^ – One sequence v. many thousands:^ • BLAST or FASTA • Some are much slower, but guarantee the“optimal alignment”^ – Smith-Waterman is the de facto standard

Dot-plots: Dotlet

http://myhits.isb-sib.ch/cgi-bin/dotlet Example: In Archaeal Genome browser, bring up

Pyrobaculum aerophilum Select CRISPR2 region (chr:45,423-46,754) to compare to CRISPR6-7region (chr:1,898,656-1,899,678) Get DNA, paste into Dotlet one at a time, giving descriptive labels,Zoom 1:5, Are there direct or inverted repeats in each CRISPR (against itself?) Relative to each other, are these direct or inverted repeats?

Assessing Alignment Significance Most Basic Rules of thumb: Two nucleotide sequences – at least 70%identical, they are likely homologous Two protein sequences – at least 25% identicalover 100 amino acid alignment Does not take into account precise length ofalignment, or number of gaps! Not sufficient to quantitatively rank hits from adatabase search

The “Twilight Zone” • Less than^ 25% sequence identity for twoprotein sequences • May still be homologous, but only similarityof 3-D protein structures can verify similarfunction (structural comparison tools todetect these discussed later in quarter) • Must have a good / near optimal alignmentfor most distantly related proteins

• Dynamic Programming • Fancy term for type of algorithm used to get the“optimal” or best possible alignment between twosequences • Needleman and Wunsch (1970) most basic method – Gives the “global” (end to end) best alignment • Smith-Waterman based closely on this algorithm, butallows for “local” alignments (best subsequencematch only) • See simple example of Global v. Local alignments inbook, Figure 3.1 p.

Basic Example

• Find best global alignment of twosequences:^ G^ A^

T^ C G T G^ C

Which is better? Match +1, Mismatch –1, Gap -

G^ A^ T^ C^

|^ |^

OR^ (Score

=^ 0)

G^ T^ G^ C G^ A^ T^ -^

C^ +1-1+1-1+

|^ |^

|^

(Score^ =

G^ -^ T^ G^

C

Moral: Scoring Model Matters!! • For DNA, model can be very simple: • +1 match, -1 mismatch • However, not all mutations have equallikelihood: • Transition: A<–>G

or^ C <–> T

• more likely • Transversion: A<–>C

or^ G <–> T

-^ less likely

Same Matrix (*10) A C^ G^

T

A^6

C^1

G^2

T^1

Actual values not important, only values relative toeach other

Protein Matrices, Same Idea • Original: Dayhoff matrix aka PAM • PAM = Percent accepted mutations • Based on small number of

correctly^ aligned proteins • Simply count how often each amino acid issubstituted for another • Frequency of substitutions based on propertiesof amino acids relative to each other