Machine leanring algorithms for single-cell data analysis
Peng, Jianhao
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/115567
Description
Title
Machine leanring algorithms for single-cell data analysis
Author(s)
Peng, Jianhao
Issue Date
2022-04-22
Director of Research (if dissertation) or Advisor (if thesis)
Milenkovic, Olgica
Doctoral Committee Chair(s)
Milenkovic, Olgica
Committee Member(s)
Ochoa, Idoia
Raginsky, Maxim
Shormonoy, Ilan
Department of Study
Electrical & Computer Eng
Discipline
Electrical & Computer Engr
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
single cell
online algorithm
network analysis
gene regulatory network
Abstract
In this thesis, we proposed various machine learning algorithms for analyzing different types of single cell sequencing data. Starting with the most common single cell RNA-seq data in Chapter 2, we proposed an online convex matrix factorization algorithm named online cvxMF that can efficiently learn representatives and interpretable lower-dimension basis vectors for each cell type. In Chapter 3, we introduced ChIA-Drop, a new type of network-structured data for chromatin interaction analysis, and extended our online cvxMF algorithm to a novel online convex network dictionary learning method that includes MCMC sampling and Gene Ontology enrichment analysis. The newly proposed method, online cvxNDL, is able to accurately reconstruct the original ChIA-Drop network and provide network dictionaries associated with biological functions. Lastly in Chapter 4, we proposed SimiC, a single cell gene regulatory network (GRN) inference algorithm that can jointly learn several GRNs from related cell phenotypes. Combined with regulon activity scores and regulatory dissimilarity scores for each of the driver genes across different phenotypes, SimiC is able to capture regulatory dynamics that are missed by previous methods.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.