Model-population analysis and its applications in chemical and biological modeling

Model-population analysis (MPA) was recently proposed as a general framework for designing new types of chemometrics and bioinformatics algorithms, and it has found promising applications in chemistry and biology. The goal of MPA is to extract useful information from complex analytical systems, so as to lead to better understanding and better modeling of chemical and biological data.To give an overall picture of MPA, we first review its key elements. Then, we describe the theories and the applications of selected methods that focus on the two fundamental aspects in chemical and biological modeling: outlier detection and variable selection. We highlight the key common principles of these methods and pinpoint the critical differences underlying each method.

► Model-population analysis is a framework for chemometrics/bioinformatics algorithms.
► We address model-population analysis (MPA) approaches to outlier detection.
► Model-population analysis is used in variable selection/biomarker discovery.