Computational Tools to Expand the Coverage of Metabolite Profiling via LC-MS/MS

<p>Metabolite profiling is a recent addition to the suite of high-throughput phenotyping technologies. Unlike measuring gene and protein expression, it is often difficult to anticipate all of the metabolites that are likely to be present in a particular sample. Still, recent efforts to catalog all present metabolites have turned up fewer metabolites than expected. This thesis focuses on one of the common metabolite profiling technologies, LC-MS/MS, and takes two steps to expand the coverage of current metabolite profiling experiments.</p><p>First, a method to enable collection of MS/MS data from a greater portion of the detected ions is presented. This is done by collecting fragmentation spectra with less stringent ion isolation settings, which provides greater sensitivity at the expense of obtaining fragments from contaminating ions. Those contaminated spectra are then computationally deconvolved, resulting in fragmentation data from ions that would have previously not been considered sufficiently abundant for fragmentation.</p><p>Second, the design of large-scale triple quadrupole metabolite profiling experiments using the METLIN metabolite database is explored. Likelihoods of obtaining common fragmentation products from each structure in the database are estimated, and are then used to identify ensembles of the most informative fragmentation spectra. The results are encouraging and provide a basis for interpretation of precursor-ion and neutral-loss scanning experiments, as well as a foundation for designing large scale multiple reaction monitoring experiments.</p><p>With these two additions to the arsenal of LC-MS/MS based metabolite profiling tools, we move closer to one day reaching the goal of global metabolite profiling. In turn, this may lead to exciting discoveries that uncover unknown pockets of human metabolism.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Biology and Biomedical Sciences: Computational and Systems Biology

Year dc:date.available

2014

Author dc:creator

• Nikolskiy, Igor

Contributors dc:contributor

• Gary J Patti

Language dc:language

English (en)

DOI

https://doi.org/10.7936/K72805MR

OAI identifier oai:identifier

oai:openscholarship.wustl.edu:etd-2384