1. Comparative Motif Finding
CS 374 – Lecture 23
Mayukh Bhaowal

2. Reference Papers
Xiaohui Xie, Jun Lu, E. J. Kulbokas, Todd R. Golub, Vamsi Mootha, Kerstin Lindblad-Toh, Eric S. Lander, Manolis Kellis, “Systematic discovery of regulatory motifs in human promoters and 3’UTRs by comparison of several mammals”, Nature, 2005
Mathieu Blanchette and Martin Tompa, “Discovery of Regulatory Elements by a Computational Method for Phylogenetic Footprinting”, Genome Res :

3. What is a Motif ?
A motif is a nucleotide sequence pattern and has biological significance.
Regulatory motifs are DNA fragments

4. Motif Logos
Height of letters represents probability of being found in that location in the motif

5. Why is it difficult to find them?
1. Short fragments
2. Degenerate
3. Unpredictable
Motifs can occur in either strands.

6. Promoter
In genetics, a promoter is a DNA sequence that enables a gene to be transcribed. The promoter is recognized by RNA polymerase, which then initiates transcription.

7. 3’ UTR
The three prime untranslated region (3' UTR) is a particular section of messenger RNA (mRNA).
An mRNA codes for a protein through translation. The mRNA also contains regions that are not translated. In eukaryotes the 5' untranslated region, 3' untranslated region, cap and polyA tail.
Image source :

8. What the paper proposes
What? Discovering the regulatory motifs in human promoters and 3’ UTRs.
How? By comparing sequence motifs of several mammals. That’s why it is called comparative motif finding.
Which mammals? Human, mouse, rat, dog.

9. Conservation properties

10. Methods Type Total Sequnce Promoter 68 Mb 3’ UTR 15 Mb Intron Control
Chose 17,700 well annotated genes from RefSeq database.
Promoters = 4kb centered at transcriptional start site (only noncoding)
3-UTRs = based on annotation of reference mRNA
Intronic sequences as a control (last two introns from each gene)

11. Motif Conservation Score
A motif is said to be conserved when an exact match is found in all 4 species.
Conservation = conserved occurrences/all occurrences
MCS =
Observed conservation
– random conservation
Standard deviation

12. Known highly conserved motif
Err α [TGACCTTG]
Of the 434 times err α occurs in human promoter regions, 162 of them are conserved across all the 4 species.
Conservation rate = 37%
Random 8-mer motif shows only 6.8% conservation rate

14. Results: Promoter Region
174 highly conserved motifs (MCS > 6)
59 strong match to known motifs, 10 weaker match.
105 potential new regulatory motifs

15. Approaches to explore biological significance
So why is the motif biologically significant?
1. tissue specificity
2. positional bias

16. Tissue Specificity
Tissue specificity of expression for genes containing discovered motifs
Expression data for 75 tissues
59 of 69 known, and 53 of 105 unknown show tissue specificity

17. Position Bias Motifs show position bias
Conserved motifs show strong position bias
Preferential occurrence within 100bases of TSS

18. Results: motifs in 3’ UTRs
In UTR 106 conserved motifs found (MCS>6)
3’-UTR motifs have not studied before
Comparison of discovered motifs to a large collection of previously known motifs not possible
Two unique properties
Strand specificity
Bias towards 8-mers

19. Property1: strand specificity
Xie, X. et al., Nature, 2005

20. Property2 : bias towards 8-mers
Xie, X. et al., Nature, 2005

21. Digression: miRNA Single stranded RNA
transcribed from DNA but not translated into protein
Many mature miRNA start with U followed by a 7-base “seed” complementary to a site in the 3’ UTR of target mRNAs.
Thus many are 8 mers
microRNA that regulates insulin secretion by an NYU study published in Nature.

22. Inference
Thus we can infer many of the conserved 8-mer motifs act as binding sites for miRNA
Leads to discovery of 52% existing miRNA genes
Leads to discovery of 129 new miRNA genes

23. Phylogenetic Footprinting

24. Problem Definition (why?)
Major challenge of current genomics is to understand how gene expression is regulated.
An important step towards this understanding is the capability to identify regulatory elements.

25. What? Phylogenetic footprinting is
1. method for the discovery of regulatory elements
2. in orthologous regulatory regions
3. from multiple species.

26. Image source: http://www.biorecipes.com/Orthologues/code.html

27. Main idea
Coding sequences evolving at a slower rate than non-coding sequences cause selective pressure
Transition in a coding sequence can possibly alter the whole function of coded protein
Transition in a non-coding sequence (RE) may only change expression frequency of a gene

28. Phylogenetic Footprinting
Study orthologous non-coding DNA from species that are related (phylogenetic tree)
Differentiation:
Tree
Find one motif in many species
Well conserved = possible Regulatory Element

29. Formalization Given: phylogenetic tree T,
set of orthologous sequences at leaves of T,
length k of motif
threshold d
Problem:
Find each set S of k-mers, one k-mer from each leaf, such that the “parsimony” score of S in T is at most d.

30. Small Example Size of motif sought: k = 4 AGTCGTACGTGAC... (Human)
AGTAGACGTGCCG... (Chimp)
ACGTGAGATACGT... (Rabbit)
GAACGGAGTACGT... (Mouse)
TCGTGACGGTGAT... (Rat)
Size of motif sought: k = 4

31. Solution AGTCGTACGTGAC... AGTAGACGTGCCG... ACGTGAGATACGT...
GAACGGAGTACGT...
TCGTGACGGTGAT...
ACGG
ACGT
Parsimony score: 1 mutation

32. An Exhaustive Algorithm
Wu [s] = best parsimony score for subtree rooted at node u,
if u is labeled with string s. …
ACGG: + ACGT: 0
...
… ACGG: 1 ACGT: 0
...
AGTCGTACGTG
ACGGGACGTGC
ACGTGAGATAC
GAACGGAGTAC
TCGTGACGGTG
… ACGG: ACGT :
… ACGG: 2 ACGT: 1
...
… ACGG: ACGT :
… ACGG: ACGT : …
ACGG: 1 ACGT: 1
\... …
ACGG: 0 ACGT: 2
...
… ACGG: 0 ACGT: +
...

33. Simple Recurrence Wu [s] =  min ( Wv [t] + h(s, t) ) Words Good:
v: children t of u
Words Good:
K-mer score at a node is the sum of its
children’s best parsimony scores for that k-mer

34. Running Time Wu [s] =  min ( Wv [t] + h(s, t) )
v: children t of u
Number of
species
Average
sequence
length
Motif length
Total time O(n k (42k + l ))
O(k  42k ) time per node

35. Results
Metallothionein Gene Family
Insulin Gene Family
C-myc promoter

36. Metallothionein Gene Family
Large number of promoter sequences
Large number of RE
Binding sites occurs within 300 bp of start codon
590 bp of sequence located upstream of start codon
Conserved elements of lengths 7,8,9,10 (K values)
Identified 12 motifs of which 4 have been confirmed
Analysis

38. Insulin Gene Family two rodents and a pig (two gene copies each)
motifs with 0 mutations, K=8
motifs with 1 mutation, K=9,10
4 conserved motifs identified
Several binding sites missed as they contain very few mutations
Analysis

39. C-myc Promoter 7 species analyzed
Contains members from diverse animal phyla (fishes, birds, mammals, batrachians)
4 of 9 predictions known are binding sites
Most located in 120 bp promoter region
Analysis

40. Drawbacks
Some binding sites does not have significant matches to most other species
Some binding sites show good conservation rate in sequences shorter than footprinter looked at
T3R

41. Drawbacks cont’d Deletions/Insertions
Failure to meet statistical significance
Some TFs bind as dimers where the binding site may consist of 2 conserved regions, separated by a few variable nucleotides

42. Thank You