Question

Compare and contrast the phylogenetic methods known as phenetics and cladistics in terms of how ancestral relations are estimated (inferred). Which one focuses on numerical comparisons and which one tries real hard to eliminate homoplasy? Briefly describe a modern algorithm that is used to generate phylogenetic trees.

Answer 1

Phenetics

Phenetics is the area of study that classifies organisms based on their similarity. It may include classifying organisms based on their structure, morphology and other traits that are observable. Phenetics disregards evolutionary relationships. Therefore, it is a primitive way of classifying organisms. It is closely related to numerical taxonomy which is concerned with the use of numerical methods for taxonomic classification.

Method used in Phenetics- During phenetic classification, clustering or ordination of organisms takes place based on phenotypes that could be observed by the taxonomist or the scientist. Therefore, dozens of characters are analyzed before clustering organisms. Then the representation of these characters using graphs takes place. This leads to the grouping of the organisms. So Phenetics focuses on numerical comparisons.

The main drawback of phenetics is that there can be a misinterpretation of organisms into groups as only physical visible characters are considered. This could give false results during the classical clustering process. Therefore, in classifying, it is important to analyze many methods of grouping for accuracy, especially the phylogenetic data of organisms.

Cladistics

Cladistics is the study area that classifies organisms based on their ancestry. Thus, cladistics considers evolutionary relationships. In cladistics, the most common recent ancestor of a particular organism plays a key role in the classification of that organism. Therefore, cladistics depends on evolutionary relationships of organisms since the similarity between organisms relies on evolutionary patterns.

Method used in Cladistics- In generating a classification of organisms based on cladistics, a cladogram is used. The cladogram is a tree-shaped diagram. Initially, during the generation of cladograms, physical and morphological characteristics were used. However, at present, cladistics depend mainly on genetic sequences and phylogeny. Therefore, cladograms are also referred to as phylogenetic trees at present. Furthermore, at present, cladistics uses morphological, evolutionary and phylogenetic data in classification.

The branches of the cladograms define various distances of similarity between the organisms and the deviation of organisms. The accuracy of classifying organisms based on cladistics is more accurate and efficient. Bioinformatics plays a major role in cladistics, increasing the accuracy and the acceptability of the results. Therefore Cladistics tries real hard to eliminate homoplasy.

Summary

So the key differences between phenetics and cladistics relies on the characteristics each method considers in classification. Thus, phenetics considers only structural and morphological characteristics, while cladistics considers evolutionary and ancestry characters. Due to this reason, the accuracy and reliability of the two techniques also vary. Moreover, the clustering process also differs in the two methods.

The below infographic presents more information regarding the difference between phenetics and cladistics.

There are several algorithms used to infer phylogenetic trees, but the most widely-used algorithms fall into three main categories:

1. Distance algorithms use a relevant metric of distance (Hamming distance, 1 - sequence identity, expression level, etc.) to construct a matrix showing the pairwise distance between all organisms/genes in a dataset. This matrix is then used to construct a tree, with organisms/genes appearing on the terminal nodes (tips), such that the distance between tips in the tree is equal to (or at least close to) the distance between organisms/genes in the matrix. An example of this is the neighbor-joining algorithm.
2. Maximum parsimony algorithms attempt to find trees with the highest parsimony score for a given data set. If the data set is a sequence alignment, then the maximum parsimony tree is the one that allows us to produce all of the sequences in the alignment from a single ancestral sequence at the root of said tree using the smallest number of single mutations. (Importantly, there can be multiple maximum parsimony trees for the same data set.) A disadvantage of this method is that it is inconsistent under certain conditions, namely, when comparing organisms/genes from highly divergent lineages (a problem known as “long branch attraction”).
3. Likelihood algorithms come in two distinct flavors: 1) maximum likelihood and 2) Bayesian inference. Both approaches involve computing the “likelihood” of phylogenies (i.e., the probability of observing your sequence alignment given that the sequences evolved along a specific tree according to a model of evolution), and they are used almost exclusively on molecular data sets.
   • Maximum likelihood (ML) algorithms typically employ a “hill climbing” strategy.
   • Bayesian inference (BI) programs use a Markov chain Monte-Carlo (MCMC) algorithm to estimate the posterior distribution of phylogenies by sampling trees in proportion to their likelihood.