Molecular Evolution and Phylogeny

1
Molecular Evolution and Phylogeny

• Examples

2
Weakly deleterious mutations

• Weakly deleterious mutations can reach high
  frequencies in local populations and, thus, may
  contribute significantly to genetic variance in
  disease susceptibility.

3
Sequencing of human polymorphisms

• A team at Celera Genomics sequenced by
  exon-specific polymerase chain reaction (PCR)
  amplification 20,362 loci in 20 European
  Americans, 19 African Americans and one male
  chimpanzee with the initial intention of finding
  novel nonsynonymous single nucleotide
  polymorphisms (SNPs) based on their 2001 build of
  the human genome.

4
Divergence between human and mouse

• A total of 34,099 fixed synonymous differences
  between 39 humans and the chimpanzee yield a
  genomic average synonymous divergence of dS
  1.02.
• 20,467 non-synonymous differences dN 0. 242
  across 11.81 megabases (Mb) of aligned coding DNA.

5
Polymorphisms

• 15,750 synonymous and 14,311 non-synonymous SNPs
  among the human subjects, yielding average
  synonymous and non-synonymous SNP densities of pS
  0.470 and pN 0.169.

6
Polymorphisms are more than divergence

• a highly significant excess of amino acid
  variation relative to divergence.

7
Can you comment on the following?

• Evolution of human populations since sharing a
  last common ancestor with chimps
• Type of nonsynonymous mutations (very deleterious
  or mildly deleterious) in human populations
• Positively selection
• Negative selection
• Disease associations?

8
Non-neutral evolution

• dN/dS 1 neutral evolution
• dN/dS gt 1 positive selection
• dN/dS lt1 negative selection

9
Accelerated evolution of genes
10
What makes us a vertebrate?

• Neural crest?
• Highly sophisticated nervous system?
• Bones/cartilage?
• Vertebrate specific genes?

11
Origin of bilateria

• Some vertebrate genes date prior to the origin of
  bilateria

12
Bilateria

• Bilateria a monophyletic group of metazoan
  animals characterized by bilateral symmetry.

13
Radial symmetry

• Bilateria excludes the Cnidaria, Ctenophora (sea
  gooseberries), Porifera (sponges) and Placozoa.

14
A little taxonomy
15
Cnidaria

• Cnidaria a basal phylum, has two body layers,
  radial symmetry and being at the tissue grade of
  morphological organization.
• There are two basic morphologies the sessile
  polyp and the swimming medusa or jellyfish.
• The phylum contains four classes (examples),
  including jellyfish, sea anemone and hydra

16
Body Axis

• Oralaboral axis the single obvious body axis of
  the two radiate phyla (Cnidaria and
  Ctenophora), marked at one end by the mouth or
  oral pore.

17
Wnts signaling
http//www.stanford.edu/rnusse/reviews/NaVReviewF
inal438747a.pdf
18
Wnt Signaling

• In Wnt signalling pathway, ligand binding
  triggers the formation of a receptor complex, and
  protein kinases modify the receptor tails,
  leading to recruitment of cytoplasmic factors.
• In other signalling pathways, receptor-induced
  protein phosphorylation amplifies the signal, and
  the receptor-associated kinase acts as a catalyst
  for the modification of many substrate molecules.

19
Wnt genes

• Mammals have 19 wnts
• Sea anemone has 12
• Nematostella vectensis, a diploblast

Kusserow A, Pang K, Sturm C, Hrouda M, Lentfer J,
Schmidt HA, Technau U, von Haeseler A, Hobayer B,
Martindale MQ, Holstein TW (2005) Unexpected
complexity of the Wnt gene family in a sea
anemone. Nature 433156-160.
20
Nematostella vectensis
http//www.nematostella.org/
21
Phylogenetic tree of wnts
22
Expression of wnts
The original bilaterian was equipped with a
fairly elaborate set of molecular tools.
23
Endoderm, ectoderm, mesoderm

• For example, the Nematostella ectodermal genes,
  NvWnt1, NvWnt2, NvWnt4 and NvWnt7 correspond to
  the neuroectodermal Wnt genes in the higher
  Bilateria.
• NvWnt5, NvWnt6 and NvWnt8 are expressed in the
  endoderm, whereas the corresponding genes in
  deuterostomes are all expressed in the mesoderm.

24
Collagen

• Bone is significantly linked to cartilage, both
  in development and evolution, with earlier forms
  having a cartilaginous skeleton that is replaced
  by bone. In vertebrates, cartilage also contains
  threads of collagen running through it.

25
Collagen

• Bone is a living tissue continually remodeling
  the mineral matrix threaded with fibers of a
  protein, type II collagen, gives strength.

26
Collagen

• Collagen is an ancient protein (800 million years
  ago?).
• There are about 27 different types of collage in
  at least a dozen different classes.
• http//web.indstate.edu/thcme/mwking/extracellular
  matrix.html
• One particular type, type II collagen, is an
  essential part of the matrix of bones and
  cartilages.

27
A primitive jawless fish from the late Devonian,
around 370 million years ago. Do lampreys have
collagen?
28
Initially it was thought lampreys dont have
collagen

• Zhang et al. screened a library of lamprey
  sequences and isolated two forms of collagen II,
  Col2a1a and Col2a1b.
• The presence of a collagen homolog related to
  human collagen II the gene arose before the
  (jawless)lamprey-gnathostome (true-jaws) split.
• Col2a1 is used in developing branchial
  cartilaginous skeleton.

Proc Natl Acad Sci U S A. 2006 Feb 21 Lamprey
type II collagen and Sox9 reveal an ancient
origin of the vertebrate collagenous
skeleton. Zhang G, Miyamoto MM, Cohn MJ.
29
but they do!
30
Collagen phylogeny
31
Bootstrapping

• The bootstrap is a procedure that involves
  choosing random samples with replacement from a
  data set and analyzing each sample the same way.

32
Bootstrapping

• Sampling with replacement means that every sample
  is returned to the data set after sampling. So a
  particular data point from the original data set
  could appear multiple times in a given bootstrap
  sample.

33
Bootstrapping

• The number of elements in each bootstrap sample
  equals the number of elements in the original
  data set. The range of sample estimates we obtain
  allows us to establish the uncertainty of the
  quantity we are estimating.

34
Reliability of a tree

• reliability of an estimated tree is to examine
  the reliability of each interior branch.

35
Bootstrap

• the reliability of an inferred tree is examined
  by using Efrons bootstrap resampling technique.
• A set of nucleotide sites is randomly sampled
  with replacement from the original set, and this
  random set is used for constructing a new
  phylogenetic tree.
• This process is repeated many times, and the
  proportion of replications in which a given
  sequence cluster appears is computed.
• If this proportion (PB) is high (say, PB gt 095)
  for a sequence cluster, this cluster is
  considered to be statistically significant.

36
Bootstrap values
37
Bootstrapping

• Open Matlab
• Open Help
• Type bootstrap and read

38
Example

• gt load lawdata
• gt plot(lsat,gpa,'')
• gt lsline

39
Plot of lsat vs. gpa
40
Calculate correlation between lsat and gpa

• gt rhohat corrcoef(lsat,gpa)
• gt rhohat
• 1.0000 0.7764
• 0.7764 1.0000

41
Is 0.78 significant?

• Now we have a number, 0.7764, describing the
  positive connection between LSAT and GPA, but
  though 0.7764 may seem large, we still do not
  know if it is statistically significant.

42
Bootstrp function

• Using the bootstrp function we can resample the
  lsat and gpa vectors as many times as we like and
  consider the variation in the resulting
  correlation coefficients.

43
Generate 1000 lsat and gpa vectors by resampling
from the original vectors

• rhos1000 bootstrp(1000,'corrcoef',lsat,gpa)
• hist(rhos1000(,2),30)

44
What is the uncertainty associated with the
observed correlation?

• gtgt mean(rhos1000(,2))
• ans
• 0.7711
• gtgt std(rhos1000(,2))
• ans
• 0.1350
• gtgt 0.13501.96
• ans
• 0.2646
• Mean /-1.96std

45
You have data on the expression pattern of two
genes

• HOXA1 and CDK6 expression values in different
  tissues are collected.
• Open the excel file named data.xls
• Copy and paste the numerial data columns (two of
  them) into the workspace as follows naming the
  data as a
• gtgt a paste.here and close bracket

46
Calculate the uncertainty associated with the
correlation btw HOXA1 and CDK6 genes

• Plot the expression values (x, HOXA1 and y,
  CDK6).
• Place a lsline on the data
• Calculate the correlation coefficient between the
  genes
• Generate 1000 bootstrapped samples to estimate
  the sample correlation coefficient.
• Determine the 95 confidence interval around the
  bootstrapped correlation coefficient.

47
Bootstrap of align2.m

• Generate 1000 samples of bootstraped alignment
  score and its 95 confidence interval using the
  bootstrp function.

48
Bayesian Inference

• There are three basic methods that have been used
  to estimate phylogeny, including distance,
  maximum parsimony (MP),and maximum likelihood
  (ML).
• Bayesian statistics differs in that in addition
  to the current data, prior knowledge is included
  in the testing of the hypothesis.

49
Medical tests and Bayesian Stats

• Assume that previous studies have evaluated the
  accuracy of this test and have shown that, if you
  are in fact ill, there is a 99 likelihood that
  the test will give a true positive result (and
  thus, a 1 likelihood that the test will give a
  false negative).

50
Medical tests and Bayesian Stats

• It was also found that if you are healthy, there
  is a 0.1 likelihood of a false positive result
  from the test. If we were simply using the data
  (i.e., the test result), we would then conclude
  that a positive test result had approximately a
  99 chance of being correct.

51
Medical tests and Bayesian Stats

• If we were to examine this question in a Bayesian
  framework, we could incorporate prior
  knowledgein this case that other studies have
  shown that the base rate of this illness is 0.1
  in the population.
• Thus, of a population of 100,000 individuals, 100
  would be ill and 99,900 would be healthy.

52
Medical tests and Bayesian Stats

• Using the likelihood values mentioned above, we
  could conclude that a positive test result would
  be seen in 99 of the ill individuals (99 true
  positives) and 0.1 of the healthy individuals
  (approximately 100 false positives).

53
Medical tests and Bayesian Stats

• This leaves us with a conclusion that if a person
  has a positive test result, there is a 99/199 or
  approximately 50 chance that the test is correct
  and this person is actually ill. Therefore, by
  including prior knowledge of the base rate of the
  illness in the population, the perceived chance
  that a positive result indicates that an
  individual actually has the illness drops from
  99 to 50.