MALA

MALA performs clustering and classification of microarray gene expression data to reduce the volume of data to be analyzed and to extract discriminative logic formulas that characterize experimental classes.

Key Features:

• Discretization: Converts continuous rational gene expression values into a limited number of intervals for each cell in the microarray.
• Gene Clustering: Groups genes by similar expression patterns to simplify the dataset.
• Feature Selection: Selects a small subset of clustered genes that exhibit strong discriminating power across classes.
• Formulas Computation: Employs optimization algorithms for feature selection and logic formula extraction to identify networks of genes characterizing classes.
• Classification: Classifies microarray experiments based on the computed logic formulas.

Scientific Applications:

• Microarray experiment classification: Assigns experimental samples to classes using logic formulas derived from gene expression data.
• Discriminative gene-network identification: Identifies compact logic formulas that represent networks of genes distinguishing classes.
• Case studies: Applied to real datasets such as comparisons of Alzheimer diseased versus control mice microarray probes.

Methodology:

Uses a machine learning process-based methodology including discretization of continuous gene expression data into intervals per microarray cell; gene clustering by expression patterns; selection of a small discriminative subset of genes; optimization algorithms for feature selection and logic formula extraction; and classification of experiments based on the computed formulas.