A Classification-based Quantitative Approach for SILAC Data

Abstract

A practical and powerful approach for stable isotope labeling is stable isotope labeling by amino acids in cell culture (SILAC). A key advantage of SILAC is the ability to detecting simultaneously the isotopically labeled peptides in a single instrument run and so guarantees relative quantitation for a large number of peptides without introducing any variation caused by separate experiments. In this work, we introduce a new quantitative approach to dealing with SILAC protein-level summary using classification-based methodologies. Unlike existing methods, our approach depends mainly on the protein ratio summary and is not restricted only to the proteins with two or more peptide hits. In particular, our approach uses Gaussian mixture model and a stochastic, metaheuristic global optimization algorithm, particle swarm optimization (PSO), to avoid either a premature of convergence or being stuck in a local optimum. Our simulation studies show that the proposed method performs the best in terms of F1 score.