Please use this identifier to cite or link to this item: http://hdl.handle.net/10125/63211

Building a large-scale single-cell multi-omics analysis platform for bench scientists

File Size Format  
Zhu_hawii_0085A_10274.pdf 11.71 MB Adobe PDF View/Open

Item Summary

Title:Building a large-scale single-cell multi-omics analysis platform for bench scientists
Authors:Zhu, Xun
Contributors:Okimoto, Gordon (advisor)
Molecular Biosciences and Bioengineering (department)
Keywords:Bioinformatics
Computer engineering
Analysis platform
Multi-omics
Single-cell
Date Issued:2019
Publisher:University of Hawai'i at Manoa
Abstract:With the astonishing advances in Single-cell omics technologies, the integration of different omics analysis has been predicted to be the next big step. However, there has not been a dedicated platform enabling this integration. In this thesis, my colleagues and I developed an online platform that allows for analyzing single-cell data and multi-omics data. This platform, nick-named “Granatum”, works to bridge the gap between the single-cell bioinformaticians who develop advanced computational methods and the single-cell bench scientists who design and conduct the experiments but lack the programming experience to utilize these methods. The goal is to help form a Granatum-connected community where the latest methods and tools developed by bioinformaticians can be quickly shared, organized, and packaged in a user-friendly manner. The entire project was broken down into three steps. The first step is to develop a single-cell specific analysis method. For this step I developed the NMFEM algorithm which applies the non-negative matrix factorization (NMF) technique commonly used in engineering to finding the subclusters and important genes in RNA-Seq data. I combined NMF with the spin-glass community detection algorithm to discover biologically interesting pathways or function groups among the NMF-selected genes. I tested the algorithm on five different mouse and human scRNA-Seq datasets and concluded that NMFEM outperforms traditional clustering methods in its accuracy.
The second step is to develop an online analysis platform specifically for scRNA-Seq. For this step, I developed the Granatum platform using the GUI framework Shiny, in the R programming language. The platform graphically guides the user through the parsing, preprocessing, and analysis of scRNA-Seq data via an 1interactive, graphical user interface. The platform was released online and it has so far gained thousands of users and overall positive user feedback.
The third and last step is to develop an integrative platform for joint analysis of multi-omics data. For this step, I first redesigned the Granatum platform from scratch to allow for many improvements in response to the user feedback. The new platform has a modularized architecture and has a centralized app store which hosts the individual analysis modules called Gboxes. I then implemented the JAMMIT multi-omics algorithm on Granatum with my unique two-step scanning/analyzing approach and demonstrated the correctness and effectiveness of my implementation with multi-omics data downloaded from TCGA.
Description:Ph.D. Thesis. Ph.D. Thesis. University of Hawaiʻi at Mānoa 2019
Pages/Duration:154 pages
URI:http://hdl.handle.net/10125/63211
Rights:All UHM dissertations and theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission from the copyright owner.
Appears in Collections: Ph.D. - Molecular Biosciences and Bioengineering


Please email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.

Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.