Presentation is loading. Please wait.

Presentation is loading. Please wait.

Tree Building ß What is a tree ?  Cladograms  Trees  Scenario ß How to build a tree ?  Observations  First Principles  Assumptions  Methods.

Similar presentations

Presentation on theme: "Tree Building ß What is a tree ?  Cladograms  Trees  Scenario ß How to build a tree ?  Observations  First Principles  Assumptions  Methods."— Presentation transcript:

1 Tree Building ß What is a tree ?  Cladograms  Trees  Scenario ß How to build a tree ?  Observations  First Principles  Assumptions  Methods

2 What is a tree ? ß Cladograms and Trees  Both are graphs in mathematical terms: A graph is a collection of nodes (vertices) and lines / branches (edges) connecting the nodes. A cladogram/tree for our purposes is allowed at most one edge between any two vertices.

3 Cladogram & Trees - 1 ß The degree of a node is the number of branches that contain that node. ß A node of degree 1 is called a leaf (or terminal node) ß All nodes that are not leaves are called internal.

4 Cladograms & Trees - 2 ß A tree is elementary if no node has degree 2 (‘net-work’ in cladistic jargon). ß A root is a distinguished node with degree 2, locating the ‘start’ of the tree.

5 Cladograms & Trees - 3 ß An unrooted tree is binary if every node has degree 1 or 3. ß A rooted tree is binary if it has a root of degree 2 and every other node has degree 1 or 3.

6 Cladograms & Trees - 4 Leaf Root Node Branch CAB Label ? ß Labeled rooted binary tree

7 Cladograms & Trees - 5 ß What’s the difference?  Cladogram Cladogenesis: branching events as indicated by character state changes  Tree + Anagenesis: amount and duration of change + inference of ancestor- descendant relationships

8 A Cladogram is a: ß Statement about the distribution of (shared) character states. ß Branching diagram depicting nested sets of synapomorphies resulting in a summary statement of sister- group relations among taxa.

9 Nested sets of Synapomorphies ß Detection of relationships by distribution of character-states in species X, Y, and Z. abcde aBCde aBCdEabcdE BC Syn- apomorphy E convergence ad Sym- plesiomorphy XZY

10 Relationship, and Kind of Groups ß Relationship criterion:  Recency of common ancestry ß “A species X is more closely related to another species Y than it is to another species Z if, and only if, it has at least one stem species in common with species Y that is not a stem species of Z” (Hennig, 1966, p.74) ß X and Y are sistergroups.

11 A Phylogenetic Tree is a: ß Branching diagram where:  the nodes represent real or hypothetical ancestors,  the branching represents speciation, and  the branches represent descent with modification.

12 Cladograms & Trees - 6 ß Cladogram = set of trees Every picture tells a story ABC = ?

13 Cladogram = Set of Trees C B A B C A ABC = ? ? A C B ? A BC ? A BCBA C ? B AC ?

14 The Cladistic Party Line... ß “There is simply no possible way to distinguish ancestors from extinct lineages.” ß “… if something is in fact an ancestor, there are no data that can refute the hypothesis that it is an extinct lineage and not an ancestor.” (Mark Siddal, 09/01/96,

15 … and its Counterpart ß “…’A is the ancestor of B’ is a perfectly valid hypothesis, and one that is easily falsified. All it would take to falsify it is to find an autapomorphy in A that is not found in B.” (Ron DeBry, 18/01/96,

16 How to build a cladogram ß Observations Character-state distri-butions over taxa (data matrix), or derivation thereof (distance matrix) ß First principles ß Assumptions Process Model Data Type and Quality

17 First Principles ß Evolution (descent with modification) occurs. ß Evolution results predominantly in a hierarchical scheme of relationships among the entities involved. ß...?

18 Assumptions - 1 ß “ The fact that parsimony methods are known to fail in reconstructing phylogeny when there are unequal rates of evolution, and fail in a systematic way (e.g., put long branches together when they really should each go with one of the short branches) suggest […] that certain conditions of the process of evolution have to be met in order for the method to be useful […]. If a method is only useful when certain conditions of the evolutionary process are met, I would think that these conditions might as well be thought of as assumptions.”  (Andrew J. Roger, 08/01/96,

19 Process ß “I have a pretty good idea of how evolution works, thus I can check how my data fit these ideas.” ß “Given the phylogeny, what is the probability to find the data as I did ?” ß Model = Statistical Framework  Maximum likelihood

20 Assumptions - 2 ß “The philosophical part that deserves more explanation is how you get from whatever general principles you invoke (‘parsimony’) to the specific numerical method used.” ß “Compatibility methods represent discarding a character because it has some sign of conflict with others. If there are two kinds of characters, really horribly noisy and pretty clean, that is a sensible thing to do. If there are instead two kinds, pretty clean and a little noisy, it is not. So I do not see how the principle of parsimony decides in advance which of these situations we are facing.”  (Joe Felsenstein, 14/12/95,

21 Data ß “My data will tell me what the optimal set of branching events is and from there I will try to grasp what actually could have happened.”  Parsimony  Group / Component Compatibility  Character Compatibility

22 Observations ß Molecular data  DNA sequences: nuclear, mitochondrial, ribosomal  DNA-DNA hybridization  Restriction-site and -fragment  Allelic isozymes ß Morphological data ß Anatomical data ß Chemical data

23 Principles - 1 Black Boxes ? Observations Methods Assumptions Phylogenetic Trees Cladogram(s) Assumptions Optimality

Download ppt "Tree Building ß What is a tree ?  Cladograms  Trees  Scenario ß How to build a tree ?  Observations  First Principles  Assumptions  Methods."

Similar presentations

Ads by Google