Geometry of single-cell multiplex data reflects biophysical processes in cells
Citation
Wang, Shu. 2021. Geometry of single-cell multiplex data reflects biophysical processes in cells. Doctoral dissertation, Harvard University Graduate School of Arts and Sciences.Abstract
Recent advances in experimental techniques allow for measuring multiple (10 - 10^5) biomolecularparameters in single cells from culture and tissue. The information-rich data obtained by these
single-cell multiplex techniques are a valuable resource for studying the multiple scales within an
organism, from the complex biomolecular signaling networks within a cell to the intricate tissue
organizations formed by cell populations. The interpretation and analysis for data of such high dimensionality
is challenging; incorporating knowledge of biophysical mechanisms, such as reaction
kinetics or inter-cellular interactions, with data analysis is yet another challenge. Single-cell multiplex
data ought to provide unique information about the relations and mutual constraints between
biophysical mechanisms across scales, if the aforementioned challenges can be overcome.
In this thesis, I analyze covariances between single-cell biomolecular abundances, and their correlation
functions over the spatial scales of cell populations, as a basic means to characterize the high-dimensional
geometry of single-cell multiplex data distributions. I then illustrate a connection between
observed data geometry and manifold geometry that arises from the mass action kinetics commonly
used to model biochemical reactions. Specifically, I derive a relation between the covariance
of biomolecular abundances and the reaction network topology of a general class of mass-action
kinetics models. I also empirically observe that the spatial correlations of biomolecular abundances
between interacting homogeneous epithelial cells obey simple constraints.
I then survey the auto- and cross-correlations of different cell-types in colorectal cancer tissues,
and statistically relate the intricate histo-morphology of tumors to the protein expression of single
cells. This reaffirms that there are strong, predictive links between biomolecular abundances and the
developmental organization of tissues.
Terms of Use
This article is made available under the terms and conditions applicable to Other Posted Material, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAACitable link to this page
https://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37368333
Collections
- FAS Theses and Dissertations [6902]
Contact administrator regarding this item (to report mistakes or request changes)