BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:iCalendar-Ruby
BEGIN:VEVENT
CATEGORIES:Defenses
DESCRIPTION:"Statistical Modeling for High-dimensional Omics Studies for co
 ngruence\, Heterogenity and Clustering" - Public Health/Biostatistics\, \n\
 nCommittee:\nGeorge Tseng (advisor and committee chair)\n\nAbstract: \n\nHi
 gh-dimensional omics data generated from high-throughput technologies captu
 re molecular intricacy and variations\, providing comprehensive insights in
 to the pathological development of human diseases. However\, statistical qu
 antification of heterogeneity and congruence can be difficult both within a
  cohort and across studies due to the high dimensionality. This dissertatio
 n focuses on methodology development for cross-species congruence analysis 
 for transcriptomic responses (Chapter 2)\, multivariate guided clustering f
 or disease subtyping (Chapter 3) and multiple clustering in omics data (Cha
 pter 4).\n\nIn Chapter 2\, we propose a congruence analysis framework for t
 ranscriptomic response analysis by developing quantitative concordance/disc
 ordance scores incorporating data variabilities and pathway-centric downstr
 eam investigation. This framework can be applied to cross-species/tissues s
 tudies to assist researchers to numerically quantify and visually identify 
 molecular mechanisms and pathway subnetworks that are best or least mimicke
 d by model organisms\, providing foundations for hypothesis generation and 
 subsequent translational decisions.\n\nIn Chapter 3\, we propose a multivar
 iate guided clustering model (mgClust) to identify homogeneous molecular su
 btypes of a complex disease that are associated to multiple disease related
  clinical variables collectively. The two main components\, disease subtypi
 ng model and multivariate clinical variable association model\, interact wi
 th each other through a latent subtyping variable. Compared with existing m
 ethods\, we show that mgClust has improved clustering and feature selection
  performance with accurate clinical variable selection through extensive si
 mulations. Application to a lung disease dataset shows its benefit in enhan
 cing interpretation and mechanistic understanding.\n\nIn Chapter 4\, we pro
 pose a model-based multiple clustering algorithm to simultaneously discover
  multiple meaningful partitions of samples. Views with heterogeneous partit
 ions are achieved by the competition of likelihoods in mixture models while
  clusters within each view are determined through the competition across in
 dividual Gaussian distributions. A relative likelihood of mixture models is
  proposed in the E-step of the Expectation-Maximization algorithm to enhanc
 e the view assignment and a tight clustering initialization is used to enco
 urage dissimilar views. Application to multiple human brain tissue datasets
  show its effectiveness in capturing multiple distinct perspectives nested 
 in high-dimensional omics data.\n\nContribution to public health: The frame
 work proposed in Chapter 2 provides a quantitative approach to identify bio
 markers\, pathways and topological gene regulatory modules that are best or
  least mimicked by the model organism\, which will facilitate hypothesis ge
 neration and translational guidance of animal models. The model proposed in
  Chapter 3 can identify disease subtypes that are associated with clinical 
 variables of interests\, which has important implication toward precision m
 edicine. Chapter 4 provides a tool for simultaneously generating multiple p
 artitions of samples reflecting different perspectives of the dataset\, fac
 ilitating the exploration of publicly available omics data and the discover
 y of new knowledge in diseases.
DTEND:20230404T190000Z
DTSTAMP:20260511T122651Z
DTSTART:20230404T180000Z
GEO:40.442859;-79.958417
LOCATION:Public Health\, 7139
SEQUENCE:0
SUMMARY:Dissertation Defense: Wei Zong
UID:tag:localist.com\,2008:EventInstance_42641823606979
URL:https://calendar.pitt.edu/event/dissertation_defense_wei_zong
END:VEVENT
END:VCALENDAR