We consider quantitative molecular data (such as gene expression, or gene methylation levels) and proceed with feature (e.g., gene or transcript) pre-selection procedure to characterize a given dataset.