(Regulatory-) Motif Finding (Regulatory-) Motif Finding

Clustering of Genes

Find binding sites responsible for

common expression patterns

Chromatin Immunoprecipitation

Microarray Chip“ChIP-chip”

Finding Regulatory Motifs

Given a collection of genes with common expression,

Find the TF-binding motif in common

.

.

.

Characteristics of Regulatory Motifs

• Tiny

• Highly Variable

• ~Constant Size Because a constant-size

transcription factor binds

• Often repeated

• Low-complexity-ish

Essentially a Multiple Local Alignment

• Find “best” multiple local alignment

• Alignment score defined differently in probabilistic/combinatorial cases

.

.

.

Algorithms

• CombinatorialCONSENSUS, TEIRESIAS, SP-STAR, others

• Probabilistic

1. Expectation Maximization:MEME

2. Gibbs Sampling: AlignACE, BioProspector

Combinatorial Approaches to Motif Finding

Discrete Formulations

Given sequences S = {x1, …, xn}

• A motif W is a consensus string w1…wK

• Find motif W* with “best” match to x1, …, xn

Definition of “best”:

d(W, xi) = min hamming dist. between W and any word in xi

d(W, S) = i d(W, xi)

Approaches

• Exhaustive Searches

• CONSENSUS

• MULTIPROFILER

• TEIRESIAS, SP-STAR, WINNOWER

Exhaustive Searches

1. Pattern-driven algorithm:

For W = AA…A to TT…T (4K possibilities)

Find d( W, S )

Report W* = argmin( d(W, S) )

Running time: O( K N 4K )

(where N = i |xi|)

Advantage: Finds provably “best” motif W

Disadvantage: Time

Exhaustive Searches

2. Sample-driven algorithm:

For W = any K-long word occurring in some xi

Find d( W, S )

Report W* = argmin( d( W, S ) )or, Report a local improvement of W*

Running time: O( K N2 )

Advantage: Time

Disadvantage: If the true motif is weak and does not occur in data

then a random motif may score better than any instance of true motif

MULTIPROFILER

• Extended sample-driven approach

Given a K-long word W, define:

Nα(W) = words W’ in S s.t. d(W,W’) α

Idea:

Assume W is occurrence of true motif W*

Will use Nα(W) to correct “errors” in W

MULTIPROFILER

Assume W differs from true motif W* in at most L positions

Define:

A wordlet G of W is a L-long pattern with blanks, differing from W L is smaller than the word length K

Example:

K = 7; L = 3

W = ACGTTGA

G = --A--CG

MULTIPROFILER

Algorithm:

For each W in S:For L = 1 to Lmax

1. Find the α-neighbors of W in S Nα(W)2. Find all “strong” L-long wordlets G in Na(W)3. For each wordlet G,

1. Modify W by the wordlet G W’2. Compute d(W’, S)

Report W* = argmin d(W’, S)

Step 1 above: Smaller motif-finding problem; Use exhaustive search

CONSENSUS

Algorithm:

Cycle 1:

For each word W in S (of fixed length!)

For each word W’ in S

Create alignment (gap free) of W, W’

Keep the C1 best alignments, A1, …, AC1

ACGGTTG , CGAACTT , GGGCTCT …

ACGCCTG , AGAACTA , GGGGTGT …

CONSENSUS

Algorithm:

Cycle t:

For each word W in S

For each alignment Aj from cycle t-1

Create alignment (gap free) of W, Aj

Keep the Cl best alignments A1, …, ACt

ACGGTTG , CGAACTT , GGGCTCT …

ACGCCTG , AGAACTA , GGGGTGT …

… … …

ACGGCTC , AGATCTT , GGCGTCT …

CONSENSUS

• C1, …, Cn are user-defined heuristic constants

N is sum of sequence lengths n is the number of sequences

Running time:

O(N2) + O(N C1) + O(N C2) + … + O(N Cn)

= O( N2 + NCtotal)

Where Ctotal = i Ci, typically O(nC), where C is a big constant

Expectation Maximization in Motif FindingExpectation Maximization in Motif Finding

All K-long wordsmotif background

Expectation Maximization

Algorithm (sketch):

1. Given genomic sequences find all K-long words

2. Assume each word is motif or background

3. Find likeliest Motif Model

Background Model

classification of words into either Motif or Background

Expectation Maximization

• Given sequences x1, …, xN,

• Find all k-long words X1,…, Xn

• Define motif model:

M = (M1,…, MK)

Mi = (Mi1,…, Mi4)(assume {A, C, G, T})

where Mij = Prob[ letter j occurs in motif position i ]

• Define background model:

B = B1, …, B4

Bi = Prob[ letter j in background sequence ]

motif background

ACGT

M1 MKM1 B

Expectation Maximization

• Define

Zi1 = { 1, if Xi is motif;

0, otherwise }

Zi2 = { 0, if Xi is motif;

1, otherwise }

• Given a word Xi = x[1]…x[k],

P[ Xi, Zi1=1 ] = M1x[1]…Mkx[k]

P[ Xi, Zi2=1 ] = (1 - ) Bx[1]…Bx[K]

Let 1 = ; 2 = (1- )

motif background

ACGT

M1 MKM1 B

1 –

Expectation Maximization

Define:

Parameter space = (M,B)

1: Motif; 2: Background

Objective:

Maximize log likelihood of model:

2

1

2

111

1

2

11

log)|(log

))|(log(),|,...(log

j jjij

n

ijiij

n

i

n

i jjijijn

ZZ

Z

XP

XPZXXP

ACGT

M1 MKM1 B

1 –

Expectation Maximization

• Maximize expected likelihood, in iteration of two steps:

Expectation:

Find expected value of log likelihood:

Maximization:

Maximize expected value over ,

)],|,...([log 1 ZXXPE n

Expectation:

Find expected value of log likelihood:

2

1

2

111

1

log][)|(log][

)],|,...([log

j jjij

n

ijiij

n

i

n

ZZ EXPE

ZXXPE

where expected values of Z can be computed as follows:

ijii

jijij Z

XPXP

XPZE *

)|()1()|(

)|(][

21

Expectation Maximization: E-step

Expectation Maximization: M-step

Maximization:

Maximize expected value over and independently

For , this is easy:

n

i

n

i

iii

NEW

n

Zxam ZZ

1 1

121

*))1log(log(arg **

• For = (M, B), define

cjk = E[ # times letter k appears in motif position j]

c0k = E[ # times letter k appears in background]• cij values are calculated easily from Z* values

It easily follows:

4

1k jk

jkNEWjk

c

cM

4

1 0

0

k k

kNEWk

c

cB

to not allow any 0’s, add pseudocounts

Expectation Maximization: M-step

Initial Parameters Matter!

Consider the following “artificial” example:

x1, …, xN contain: 212 patterns on {A, T}: A…A, A…AT,……, T… T 212 patterns on {C, G}: C…C , C…CG,…… , G…G D << 212 occurrences of 12-mer ACTGACTGACTG

Some local maxima:

½; B = ½C, ½G; Mi = ½A, ½T, i = 1,…, 12

D/2k+1; B = ¼A,¼C,¼G,¼T;

M1 = 100% A, M2= 100% C, M3 = 100% T, etc.

Overview of EM Algorithm

1. Initialize parameters = (M, B), : Try different values of from N-1/2 up to 1/(2K)

2. Repeat:

a. Expectation

b. Maximization

3. Until change in = (M, B), falls below

4. Report results for several “good”

Overview of EM Algorithm

• One iteration running time: O(NK) Usually need < N iterations for convergence, and < N starting

points. Overall complexity: unclear – typically O(N2K) - O(N3K)

• EM is a local optimization method

• Initial parameters matter

MEME: Bailey and Elkan, ISMB 1994.

M,B