Advancing Lipidomic Bioinformatic Technologies

点击翻译按钮获取中文摘要

Forty five years ago, the lipid biochemist Albert Lehninger, known to many of us for his seminal Biochemistry textbook, predicted that the ability to identify the individual “fats” or lipids in our membranes would enable a new understanding of how brain cells communicate. Forty five years later, technological advances in mass spectrometry are allowing us to address the two seemingly simple questions required to realize Lehninger’s vision: How many lipid species are there? What effect does lipid diversity have on cellular function? These are exciting days. Hundreds to thousands of lipid species can now be profiled in different subcellular membrane compartments. Basic unitary conceptions are being challenged. Diacylglycerol, commonly conceived by cell biologists as a single lipid, is now recognized to be a family of over 50 structurally distinct species each controlling different cellular processes. Yet new challenges come with this success. The emerging field of lipidomics faces four formidable roadblocks: (1) Data Processing and Lipid Identification – There are few bioinformatic tools capable of assigning molecular identities directly from the mass spectrometer; (2) Statistical Analysis – There are few means of determining whether compositional changes in membranes are statistically “significant”; (3) Pathway Analysis – We are in desperate need of tools that identify metabolic changes underlying the compositional differences detected between membranes; (4) Modeling Tools – We have little to no capacity to model rapidly the impact that altering lipid composition has on functional membrane properties. We propose here to address some of these issues by developing new bioinformatic tools for the lipidomic analysis of neuronal membrane and membrane microdomains. To address some of the obstacles facing the emerging field of lipidomics, our short term objectives (1-3 years) are to develop immediately applicable bioinformatic tools for data processing and lipid identification. Our medium term objectives (3-5 years) are to add new Bayesian statistical methods and pathway analysis capacities into these toolkits. Our long term objectives (>5 years) will be to develop new modeling approaches to assess the spatial and dynamic aspects of compositional lipid changes in membrane systems and biophysical context. An important aspect of our proposal is the training of a new breed of bioinformatician and computational biologist; one with the knowledge and skillsets to place Canadian researchers at the forefront of this exciting new field. Our industry collaborations will also link our trainees directly to industrial partners to facilitate their subsequent career success.