Bian, Yuemin
(2021)
The Research and Development of an Artificial Intelligence Integrated Fragment-Based Drug Design Platform for Small Molecule Drug Discovery.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Drug discovery is expensive. The average cost for the development of a new drug now hits $2.6 billion USD and the overall discovery process takes over 12 years to finish. Moreover, these numbers keep increasing. It is critical to think and explore efficient and effective strategies to confront the growing cost and to accelerate the discovery process. The rapid advancement in computational power and the blossom of Artificial Intelligence (AI) brought a promising solution to the field. There is increased availability of both chemical and biological data in drug discovery. The capability of dealing with large data to detect hidden patterns and to facilitate future data prediction in a time-efficient manner favored Machine Learning (ML) algorithms. One step further from the symbolic AI that uses explicit rules to maneuver knowledge, ML allows computers to solve specific tasks by learning on their own. Promising and compelling outcomes including the identification of DDR1 kinase inhibitors within 21 days using deep learning generative models may indicate that we are probably at the corner of an upcoming revolution of drug discovery in the AI era, and the good news is that we are witnessing the change. In this thesis, an AI integrated fragment-based drug design (FBDD) platform is proposed and developed as an innovative solution to the small molecule drug discovery in the big data era. The platform is constituted of three modules, (1) module of deep learning (DL) generative chemistry, (2) module of cheminformatics and computational chemistry, and (3) module of systems pharmacology network study. At module one, DL generative modeling with cutting-edge neural network architectures including a generative adversarial network (GAN) and a recurrent neural network (RNN) can realize the automated de novo molecule generative with target specificity. ML-based decision-making models can also be constructed facilitating large scale virtual screening for early-stage hit identification. At module two, in silico modeling and simulation of cheminformatics is focused to realize both structure-based and ligand-based drug design approaches. Computational chemistry methodologies are extensively integrated to develop an FBDD approach for lead identification and modification. At module three, the concept of systems pharmacology is fused to expedite the network analysis for the given small molecules. The compound-target network, target-pathway network, and target-disease network can be generated and analyzed to contribute a comprehensive understanding of the molecule of interest.
Share
Citation/Export: |
|
Social Networking: |
|
Details
Item Type: |
University of Pittsburgh ETD
|
Status: |
Unpublished |
Creators/Authors: |
|
ETD Committee: |
|
Date: |
4 January 2021 |
Date Type: |
Publication |
Defense Date: |
9 December 2022 |
Approval Date: |
4 January 2021 |
Submission Date: |
31 December 2020 |
Access Restriction: |
2 year -- Restrict access to University of Pittsburgh for a period of 2 years. |
Number of Pages: |
485 |
Institution: |
University of Pittsburgh |
Schools and Programs: |
School of Pharmacy > Pharmaceutical Sciences |
Degree: |
PhD - Doctor of Philosophy |
Thesis Type: |
Doctoral Dissertation |
Refereed: |
Yes |
Uncontrolled Keywords: |
Drug discovery, Artificial intelligence, Deep learning, Generative modeling, Cheminformatics, Systems pharmacology |
Date Deposited: |
04 Jan 2021 17:17 |
Last Modified: |
04 Jan 2023 06:15 |
URI: |
http://d-scholarship.pitt.edu/id/eprint/40132 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
![View Item View Item](/style/images/action_view.png) |
View Item |