EVE100 F24 Phylogenetic Analysis PDF

11/26/24 Phylogenies and their uses - reconstructing the history of life FK, Ch. 2, 18 Why do it? - Basic description of nature - Often interesting from an organismal perspective - Detecting evolutionary patterns Which traits are older vs. younger...

11/26/24 Phylogenies and their uses - reconstructing the history of life FK, Ch. 2, 18 Why do it? - Basic description of nature - Often interesting from an organismal perspective - Detecting evolutionary patterns Which traits are older vs. younger Which traits are correlated with one another during evolution Which traits are correlated with particular ecologies/lifestyles What are biological properties of ancestors 568 568 Phylogenetic approaches are fundamental tools of modern evolutionary biology 569 569 1  11/26/24 The basic units of phylogenetic analysis are taxa and characters Taxa can be anything from individuals to “higher” levels such as genera, phyla, etc. Characters can be anything that can be measured, from DNA sequences to highly complex traits Distinguishing characters from character states - examples 123456789 AGGGGCTTT Nucleotide 4 is a character, that has two states, G or A AGGGGCTTT AGGAGCTTT AGGAGCTTT Dorsal fin is a character that may have states defined by the number of rays Phenotypic traits may be discrete or continuous 570 570 All phylogenetic methods require correctly identifying homologous characters, traits shared by virtue of common ancestry A sequence alignment between two or more species is a hypothesis about nucleotide homology 123456789 AGGGGCTTT AGGGGCTTT Alignment is often a substantial challenge and AGGAGCTTT may significantly affect phylogenetic inference AGGAGCTTT Convergent or parallel evolution is a problem Sp 1 T Sp 2 A Sp1 and Sp3 experienced A-to-T A Sp 3 T changes Sp 4 A 571 571 2  11/26/24 Phenetics or distance-based methods provide one approach for inferring evolutionary relationships – still often used on sequence data Based on the idea that organisms that are more similar share a more recent common ancestor compared to organisms that are less similar. All characters may be utilized Relies on the assumption that evolutionary rates are homogeneous across lineages 572 572 Cladistics – branching patterns can be reconstructed based on shared derived character states, known as synapomorphies (Willi Hennig) Premise – there is only one true tree of life. Therefore, a shared derived character state can only be explained by virtue of inheritance through a common ancestor. In this world view, shared derived character states must be used and monophyletic groups are the only valid groups. monophyletic paraphyletic 573 573 3  11/26/24 Bad inferences can result from use of shared ancestral traits, known as sympleisiomorphies, rather than shared derived character states Wrong!! 574 574 Right!! Note that a shared, “ancestral” character state can also result from a reversal to the ancestral state in a lineage. A good example would be evolution of the wingless state in some insects. 575 575 4  11/26/24 FK 16.3 576 576 Methods for determining character state polarity (i.e., ancestral vs. derived) Outgroup method – the most commonly used approach Requires some knowledge of phylogeny to pick an outgroup Assumes state in the outgroup represents ancestral state, which may not be true if there has been considerable evolution in outgroup Different outgroups may give conflicting inferences on ancestral state 577 577 5  11/26/24 Fossil record may be used to determine ancestral state Effective if good fossil record, but ineffective without such a record Poor fossil record may cause incorrect inferences Not applicable to all taxa or characters 578 578 Developmental biology may be used to infer ancestral/derived states Derived states may appear later in development than ancestral states Developmental stages of a teleost Adult tails of lungfish, sturgeon, salmon Only true if derived phenotypes evolve through modifications of terminal stages of development. Must be used with caution and a good understanding of 579 development 579 6  11/26/24 If all traits are correctly identified as homologous and all states are correctly identified as shared and derived, then all characters should agree with the true phylogeny. In reality, homoplasious characters (due to parallel and convergent evolution) can lead to conflicting patterns (i.e., phylogenetic trees) The principle of parsimony suggests that the tree with the fewest character state changes (the tree supported by the greatest number of characters) should be favored Not an inherently statistical notion. Assumes that evolution is infrequent. 0 0 1 0 1 0 0 1 0 A B C A C B B C A ancestral state = 1 1 1 580 580 The three trees represent the three possible relationships among species 1, 2, and 3. Tree A would be favored under the parsimony criterion 581 581 7  11/26/24 FK 16.2 582 582 Newer methods of phylogenetic inference (EVE 103) Maximum likelihood – instead of minimizing evolution, these methods ask about the probability of observing the data given a model and a hypothesis of a phylogeny. Often used for molecular data. Bayesian – provides probabilities of nodes (trees) given model of evolution and assuming “prior” distributions of trees and model parameters. 583 583 8  11/26/24 Rapid radiations make phylogenetic inference difficult Hawaiian silversword alliance 584 584 Slow-evolving molecules are used to resolve deep evolutionary history while rapidly evolving molecules are used to resolve recent history Phylogenetic analysis of life 585 585 9  11/26/24 Phylogenetic analyis of human mitochondrial DNA 586 586 Incomplete lineage sorting (ILS) in recently separated populations may result in disagreement between gene trees and species trees. This phenomenon is influenced by the amount of ancestral variation and the timescale of separation. FK 16.7 In (B) and (C) the way neutral, ancestral variants drift during and after species can cause the gene tree and species tree to mismatch 587 587 10  11/26/24 Another figure from a different textbook… G/T polymorphic in ancestor 588 588 Individual instances of gene tree-species tree discordance persist through time. However, with genome scale data one can often infer the likely true species tree… FK 16.8 589 589 11  11/26/24 (A) Topological discordance in the 90% gene confidence set of 458 gene trees visualized in DensiTree (78). Dario Copetti et al. PNAS 2017;114:45:12003-12008 ©2017 by National Academy of Sciences 590 Lake Victoria cichlid radiation FK 16.9 591 591 12  11/26/24 But not always… 592 592 Comparative Method The search for general patterns of evolution Correlated evolution of particular traits Temporal patterns of trait evolution 593 593 13  11/26/24 594 594 595 Felsenstein 1985 595 14  11/26/24 Assuming each species is an independent data point is like assuming a star phylogeny But what if the phylogeny really looks like this? 596 596 597 597 15