Lecutre 23: Sequence covariation and methods comparisons in bioinformatics

Very generally, there are two types of methods evaluations commonly used in bioinformatics: * empirical: based on real biological data * theoretical: based on data simulated through some model of the biological process you’re interested in

Theoretical testing is important because you know exactly what is going on in a data set, so you know exactly what your results should be. It usually cannot match the complexity of real biological systems though, so is often considered too simplistic. Empirical testing will capture the complexity of biological systems, but you’re not aware of all of the complexities so you may derive misleading results.