Single-cell transcriptomic analysis of animals across broad evolutionary scales
Abstract/Contents
- Abstract
- The newfound ability to characterize cellular heterogeneity at single-cell resolution has encouraged broad efforts to construct transcriptomic maps of cell types from whole organisms throughout the animal kingdom. However, the deeper we delve into the more uncharted territories across the tree of life, the more challenging the data analysis and interpretation become. My thesis aims to overcome this challenge along three primary axes. First, I developed the self-assembling manifold (SAM) algorithm, an iterative soft feature selection strategy to quantify gene relevance and improve dimensionality reduction. I demonstrated its advantages over other state-of-the-art methods with novel biological findings and rigorous quantitative benchmarking. This has allowed us to identify cell types in an unsupervised, robust, and sensitive fashion, and has thus facilitated the single-cell transcriptomic analysis of a wide variety of organisms. Second, to account for the fact that cell type relationships are inherently hierarchical, I integrated nonlinear dimensionality reduction techniques with SAM to improve the unsupervised detection of cell types and subtypes in hierarchically complex datasets. This approach was enabled by my scalable implementation of kernel Principal Component Analysis using implicit linear operators. Third, I built on SAM to develop SAMap, an algorithm, first of its type, for mapping single-cell atlas datasets across evolutionarily distant species. SAMap identifies homologous cell types with shared expression programs across distant species within phyla, even in complex examples where homologous tissues emerge from distinct germ layers. Comparing species across the animal kingdom, spanning mouse to sponge, SAMap reveals ancient cell type families that likely emerged early in animal evolution. Altogether, the computational methods and analyses described in my thesis provide a roadmap for cell type discovery across the tree of life and the reconstruction of cell type evolution.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2021; ©2021 |
| Publication date | 2021; 2021 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Tarashansky, Alexander Joel |
|---|---|
| Degree supervisor | Wang, Bo, (Artificial intelligence scientist) |
| Thesis advisor | Wang, Bo, (Artificial intelligence scientist) |
| Thesis advisor | Fordyce, Polly |
| Thesis advisor | Quake, Stephen Ronald |
| Degree committee member | Fordyce, Polly |
| Degree committee member | Quake, Stephen Ronald |
| Associated with | Stanford University, Department of Bioengineering |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Alexander Tarashansky. |
|---|---|
| Note | Submitted to the Department of Bioengineering. |
| Thesis | Thesis Ph.D. Stanford University 2021. |
| Location | https://purl.stanford.edu/nn655mw6101 |
Access conditions
- Copyright
- © 2021 by Alexander Joel Tarashansky
- License
- This work is licensed under a Creative Commons Attribution 3.0 Unported license (CC BY).
Also listed in
Loading usage metrics...