Charles Explorer logo
🇬🇧

Multivariate methods in taxonomy

Class at Faculty of Science |
MB120P126

Syllabus

Comparison of phenetic and cladistic (parsimony) approach to the reconstruction of evolutionary relationships, their advantages and disadvantages; alternative approaches (maximum likelihood, Bayesian method); special features of geometrical morphometrics.

Classification of characters and their use; coefficients expressing relationships between objects or characters; genetic distances; standardisation and transformation of data.

Cluster analysis; methods of cluster analysis, their advantages and disadvantages; ties.

Ordination methods; principal components analysis (PCA), geometrical interpretation, types of PCA; principal coordinates analysis; non-metric multidimensional scaling, stress.

Discriminant analysis, requirements for data; canonical discriminant analysis; classificatory discriminant analysis, classificatory rules; stepwise discriminant analysis.

Analysis of shape, geometrical morphometrics; landmarks; Bookstein shape coordinates; Procrustes analysis; thin plate splines, deformations; shape analysis (eigenshape analysis, Fourier analysis).

Annotation

Please note that the lectures are given in Czech language. English version of the course can be requested in advance if there are at least 3 students.

The aim of the lectures is to provide basic theoretical background of the use of multivariate methods in taxonomy. Special attention is paid to principal components analysis, principal coordinates analysis, cluster analysis and various types of discriminant analysis.

Several lectures are devoted also to the methods of geometrical morphometrics (analysis of shape). From among the morphometric programs in practicals students use the SYN-TAX 2000 and R.

Lectures and practicals continue in the subject Cladistics and other methods of reconstruction of evolution - MB120P83.