PROJECT SUMMARY/ABSTRACT
Non-targeted analysis (NTA) provides a comprehensive approach to analyze environmental and biological
samples for nearly all chemicals present. Despite the recent advancements in NTA, the number of confirmed
chemicals with analytical standards remains fairly small compared to the number of detected features. There
is, thus, a need to further develop computational tools to derive more chemical structures and leverage the full
potential of HRMS. Enhancing our ability to derive more chemical structures will enable the discovery of new
industrial chemicals that humans are exposed to, especially in critical windows of development, such as
pregnancy. It will also enable the discovery of endogenously produced metabolites that may be related to
biological outcomes of importance, such as preterm birth. The objective of my proposal is to develop novel
computational methods to significantly advance our ability to analyze and interpret non-targeted analysis data
from high-resolution mass spectrometry (HRMS) and apply them to study prenatal exposures to industrial
chemicals and endogenous metabolites in a large cohort of pregnant women from Northern California. My
proposal builds on my expertise in analytical and environmental chemistry and my current postdoctoral
experience in computational chemistry and applications in human exposure. I seek additional training to
develop and apply innovative computational methods to better characterize the human exposome and in
particular the exposome of preterm birth. The contribution of my proposal will be two-fold: (1) developing novel
computational structure-prediction algorithms for HRMS datasets based on MS data and physicochemical
properties (equilibrium partition ratios between organic solvents and water, e.g., octanol/water,
chlorobenzene/water, diethyl ether/water etc.) (Aim 1) and apply them to derive potential structures for
chemical features detected in a HRMS dataset from 340 maternal and 340 matched cord blood samples to
complement the limited number of chemicals identified through MS/MS and analytical standards (Aim 2); and
(2) study the interplay between the exposome and the metabolome in preterm birth using molecular interaction
networks to visualize and compare how molecular interactions between industrial chemicals and endogenous
metabolites differ between preterm and full-term birth (Aim 3). The K99 training will expand my prior research
experience through coursework, research apprenticeship, and mentored reading, with specific training in: (1)
advanced analytical skills including -omics data analysis, machine learning, and biostatistics; (2) epidemiology,
risk assessment, human exposure to chemical stressors; and (3) human pregnancy and development. The
skills acquired during this award are critical to my long-term goal to advance computational methods to better
analyze and interpret non-targeted analysis data to support efforts to better characterize the human exposome.
This work will produce new scientific knowledge to greatly advance the understanding of the influence of
environmental exposures in the development of adverse health outcomes and in particular, preterm birth.