Multiple sequence alignment based on Larry Hunters Slides

1
Multiple sequence alignment(based on Larry
Hunters Slides)

• Generalize our pairwise alignment of sequences to
  include more than two homologous proteins.
• Looking at more than two sequences gives us much
  more information
• Which amino acids are required? correlated?
• Evolutionary/phylogenetic relationships

2
Phylogenetic Trees
Analysis of 20 samples of Cytochrome c protein
Sequences Numbers represent nucleotide
substitutions in the gene for Cytochrome c
3
Sample MSA
FOS_RAT MMFSGFNADYEASSSRCSSASPAGDSLSYYHSPA
DSFSSMGSPVNTQDFCADLSVSSANF 60 FOS_MOUSE
MMFSGFNADYEASSSRCSSASPAGDSLSYYHSPADSFSSMGSPVNTQDFC
ADLSVSSANF 60 FOS_CHICK MMYQGFAGEYEAPSSRCSSA
SPAGDSLTYYPSPADSFSSMGSPVNSQDFCTDLAVSSANF
60 FOSB_MOUSE -MFQAFPGDYDS-GSRCSS-SPSAESQ--YL
SSVDSFGSPPTAAASQE-CAGLGEMPGSF 54 FOSB_HUMAN
-MFQAFPGDYDS-GSRCSS-SPSAESQ--YLSSVDSFGSPPTAAASQE-C
AGLGEMPGSF 54 .. . .
. ... .. .. ... FOS_RAT
IPTVTAISTSPDLQWLVQPTLVSSVAPSQ-------TRAPHP
YGLPTPS-TGAYARAGVV 112 FOS_MOUSE
IPTVTAISTSPDLQWLVQPTLVSSVAPSQ-------TRAPHPYGLPTQS-
AGAYARAGMV 112 FOS_CHICK VPTVTAISTSPDLQWLVQP
TLISSVAPSQ-------NRG-HPYGVPAPAPPAAYSRPAVL
112 FOSB_MOUSE VPTVTAITTSQDLQWLVQPTLISSMAQSQG
QPLASQPPAVDPYDMPGTS----YSTPGLS 110 FOSB_HUMAN
VPTVTAITTSQDLQWLVQPTLISSMAQSQGQPLASQPPVVDPYDMPGTS
----YSTPGMS 110
... . .. ..
FOS_RAT KTMSGGRAQSIG--------------------
RRGKVEQLSPEEEEKRRIRRERNKMAAA 152 FOS_MOUSE
KTVSGGRAQSIG--------------------RRGKVEQLSPEEEEKRRI
RRERNKMAAA 152 FOS_CHICK KAP-GGRGQSIG-------
-------------RRGKVEQLSPEEEEKRRIRRERNKMAAA
151 FOSB_MOUSE AYSTGGASGSGGPSTSTTTSGPVSARPARA
RPRRPREETLTPEEEEKRRVRRERNKLAAA 170 FOSB_HUMAN
GYSSGGASGSGGPSTSGTTSGPGPARPARARPRRPREETLTPEEEEKRR
VRRERNKLAAA 170 . .
.. . . .

4
Optimal MSA

• Use Dynamic Programming?
• Optimal alignment algorithm exists, but is
  O(2nln) where n is the number of sequences and l
  is the length of the longest sequence.
• 10 sequences of length 100 take 210100101023
  operations, around 1 million years at 3GHz
• Exponential algorithms strike again.
• So, approximation approaches?

5
Progressive MSA

• Start with pairwise alignments of closely related
  sequences, and then add more distantly related
  sequences one at a time.
• Requires information (assumptions) about the
  phylogenetic relationship a priori.
• Can be estimated from all pairwise comparisons.
• Give total MSA score based on sum of pairwise
  scores
• Perhaps weighted to reduce the influence of very
  similar sequences.

6
Gaps in Progressive MSAs

• How to score gaps?
• Want to align gaps with each other over all
  sequences. A gap in a pairwise alignment that
  matches a gap in another pairwise alignment
  should cost less than introducing a totally new
  gap.
• Possible that a new gap could be made to match
  an older one by shifting around the original
  pairwise alignment, but at great computational
  cost.
• Change gap penalty near conserved domains of
  various kinds (e.g. secondary structure,
  hydrophobic regions)
• CLUSTALW http//www.ebi.ac.uk/clustalw

7
Greedy algorithms

• Progressive MSA programs make the best alignment
  of a new sequence with the existing ones they can
  at the time, and then never revisit the decision.
• Even if changing an old decision (e.g. the gaps)
  could increase the score, this approach doesn't.
• Approach is called greedy (because it takes the
  best first), and is a common way to resolve
  exponential problems.

8
Problems with progressive MSA

• Depends crucially on the quality of the pairwise
  alignments, particularly among the closest
  matches.
• No suitable resolution to the problem of gap
  penalties over multiple sequences.
• Works reasonably well for closely related
  sequences. Even then, manual adjustments are
  common.

9
Iterative MSA methods

• The idea here is to start with a reasonable
  approximation to the optimal MSA (e.g. by using a
  progressive method) and then tweaking to
  improve it.
• Various optimization techniques have been tried
  here (e.g. GAs and simulated annealing).
• Key is the scoring function for the whole MSA.
• Also, what steps to take that are likely to
  improve the score.

10
Block based methods

• Another approach to iterative methods are to
  start with short local alignments (sometimes
  called blocks) and then to reduce the problem to
  aligning the regions between the blocks
• Divide and conquer is another common CS
  approach to exponential problems.
• How to find the blocks?
• DALIGN (local alignment methods)
• DCA (divide and conquer alignments)
• Tmsa (identify patterns and use them to define
  blocks).

11
Databases of MSAs

• Once they have been calculated, they can be saved
  and shared
• Pfam database of protein families. Alignments of
  large numbers of homologous proteins.
• http//www.sanger.ac.uk/Software/Pfam/index.shtml
• TigerFam database of protein families curated
  for function, rather than homology
• http//www.tigr.org/TIGRFAMs/index.shtml

12
More web sites

• Web sites offer multiple approaches to MSA.
• Interfaces to multiple different programs
• http//searchlauncher.bcm.tmc.edu/multi-align
• http//www.techfak.uni-bielefeld.de/bcd/Curric/Mul
  Ali
• Main web-based MSA servers
• http//www.ebi.ac.uk/clustalw
• http//baboon.math.berkeley.edu/mavid/ (genomic
  seqs)
• See course website for many more listings

13
Protein motifs

• Recall that local alignments can identify similar
  regions in non-homologous proteins
• These regions (sometimes called domains) often
  have shared structure and/or function.
• Example Zinc-finger DNA binding motif

14
Zinc-finger DNA binding motif
15
Protein motifs

• How to define them?
• Consensus sequence
• Regular expression
• Profile (probability for each amino acid at each
  position)

16
ProSite consensus sequences
17
Recognizing ProSite patterns

• L14 Ribosome pattern GA-LIV(3)-x(9,10)-DNS
  -G-x(4)-FY-x(2)-NT-x(2)-V-LIV
• Some matching sequences
• GIIIACGHLIPQTNGACRTYILNDRVV
• GVLLWQPKHCSNAADGAWAWFAATAAVL
• ALIVEANIIILSISGRATTFHATSAVI
• ProSite patterns can be translated into regular
  expressions, although the bounded length patterns
  (e.g. LIV(3,5) are unwieldy to write down as
  regexps.

18
Example of ProSite

• AC-x-V-x(4)-EDThis pattern is translated as
  Ala or Cys-any-Val-any-any-any-any-any but Glu
  or Asplt A-x-ST(2)-x(0,1)-VThis pattern,
  which must be in the N-terminal of the sequence
  (lt'), is translated as Ala-any-Ser or
  Thr-Ser or Thr-(any or none)-ValltCgtThis
  pattern describes all sequences which do not
  contain any Cysteines.IIRIFHLRNIThis pattern
  describes all sequences which contain the
  subsequence 'IIRIFHLRNI'.

19
Regular expressions

• Wide use in computer science. Basis of PERL
  language (see also BioPERL).For proteins,
  a language like prosite patterns is more
  intuitive, but often equivalent.

20
Profiles

• Rather than identifying only the consensus
  (i.e. most common) amino acid at a particular
  location, we can assign a probability to each
  amino acid in each position of the domain.
• Example

1 2 3 A .1 .5 .25C .3 .1 .25 D .2
.2 .25E .4 .2 .25
21
Applying a profile

• Calculate score (probability of match) for a
  profile at each position in a sequence by
  multiplying individual probabilities. Sliding
  window
• Can transform probability to significance given
  random distribution assumption

22
Applying a profile

• Calculate score (probability of match) for a
  profile at each position in a sequence by
  multiplying individual probabilities. Sliding
  window
• Can transform probability to significance given
  random distribution assumption

23
Using motifs

• Great for annotating a sequence with no strong
  homologs.
• INTERPRO is an uniform interface to many
  different motif methods and databases
• ProSite
• Prints (fingerprints multiple motifs)
• ProDom (like Pfam, but for domains)
• SMART (mobile domains)

24
Interpro example
25
InterPro example (con't).

• Then, match the pattern to a protein database

26
How do we create motifs?

• General problem of inducing patterns from
  sequences is difficult
• Classic language result (Gold) Context-free
  grammars can not be induced from only positive
  examples
• Many patterns are compatible with any MSA. How
  to decide which constituents are required?
• In general case, we need positive examples (in
  the class) but also near misses sequences that
  are similar but not members of the class.
• Not absolutely true for protein sequences.

27
Finding Consensus Sequences

• Based on local MSAs.
• ProSite consensus built from MSA on (Amos
  Bairoch's) biological intuition, tweaked by
  calculating sensitivity and specificity of the
  patterns over SwissProt.
• True (False) positives defined by Bairoch's
  understanding.
• Not an automatable procedure!

28
Creating profiles

• Given a local MSA, creating a profile is
  straightforward.
• Calculate frequency of each amino acid at each
  position to create profile.
• What to do about zero frequencies?
• Could be sampling errors, not real zero
  probabilities.
• Zero probabilities always make zero scores!
• Regularization
• pseudocounts
• Dirichlet mixtures (blend in background
  frequencies)

29
Profile example

• MSA Counts Add 1
  pseudocount
• Profiles

1 2 3 A 2 0 1B 1 4 1 C 1 0
1 D 0 0 1
1 2 3 A 3 1 2B 2 5 2 C 2 1
2 D 1 1 2
BBB ABC ABD CBA
1 2 3 A .5 0 .25B .25 1 .25 C .25 0
.25 D 0 0 .25
1 2 3 A .37 .12 .25B .25 .63 .25 C .25
.12 .25 D .12 .12 .25
30
Feature alphabets

• Amino acids can be grouped by their
  characteristics
• Size, hydrophobicity, ionizability, etc.
• An amino acid is generally in more than one group
• Can set different regularizers (pseudocounts) for
  each different feature
• Most useful when there are multiple features
  (otherwise many amino acids get same pseudocount)