Computational Functional Analysis of Lipid Metabolic Enzymes

BAGNATO, C; ARJEN TEN HAVE; PRADOS MB; BELIGNI, MV

METHODS IN MOLECULAR BIOLOGY (CLIFTON, N.J.)

Springer

Año: 2017 vol. 1609 p. 195 - 216

1064-3745

The computational analysis of enzymes that participate in lipid metabolism has both common and uniquechallenges when compared to the whole protein universe. Some of the hurdles that interfere with the func-tional annotation of lipid metabolic enzymes that are common to other pathways include the definition ofproper starting datasets, the construction of reliable multiple sequence alignments, the definition of appropri-ate evolutionary models, and the reconstruction of phylogenetic trees with high statistical support, particu-larly for large datasets. Most enzymes that take part in lipid metabolism belong to complex superfamilies withmany members that are not involved in lipid metabolism. In addition, some enzymes that do not havesequence similarity catalyze similar or even identical reactions. Some of the challenges that, albeit not unique,are more specific to lipid metabolism refer to the high compartmentalization of the routes, the catalysis inhydrophobic environments and, related to this, the function near or in biological membranes.In this work, we provide guidelines intended to assist in the proper functional annotation of lipidmetabolic enzymes, based on previous experiences related to the phospholipase D superfamily and theannotation of the triglyceride synthesis pathway in algae. We describe a pipeline that starts with the defini-tion of an initial set of sequences to be used in similarity-based searches and ends in the reconstruction ofphylogenies. We also mention the main issues that have to be taken into consideration when using tools toanalyze subcellular localization, hydrophobicity patterns, or presence of transmembrane domains in lipidmetabolic enzymes