國立虎尾科技大學 |

An Integrated Artificial Intelligence Pipeline for Drug Discovery.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	An Integrated Artificial Intelligence Pipeline for Drug Discovery./
Author:	Deng, Jianyuan.
Description:	1 online resource (200 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
Contained By:	Dissertations Abstracts International85-09B.
Subject:	Pharmacology. -
Online resource:	click for full text (PQDT)
ISBN:	9798381970326

An Integrated Artificial Intelligence Pipeline for Drug Discovery.
Deng, Jianyuan.

An Integrated Artificial Intelligence Pipeline for Drug Discovery. - 1 online resource (200 pages)

Source: Dissertations Abstracts International, Volume: 85-09, Section: B.

Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.

Includes bibliographical references

Drug discovery is an expensive process in both time and cost with a daunting attrition rate. On average, developing a new drug costs 2.6 billion US dollars and can take more than 10 years. In the past decade, the practice of drug discovery has been undergoing radical transformations in light of the advancement in artificial intelligence (AI) as well as biomedical informatics. In this dissertation, we explore ways to build an integrated AI pipeline for drug discovery.To this end, we first conduct a survey of AI techniques in drug discovery, including model architectures and learning paradigms applicable in molecular property prediction and molecule generation. With the techniques to automate drug design, one key question persists: "what constitutes the desired properties of ideal drug candidates?" In response, we present an informatics-based approach aimed at identifying key pharmacological components underlying drug-drug interactions, which provide insights for drug design. Subsequently, we introduce a large-scale observational study designed to reveal optimal properties underlying opioid analgesics with reduced overdose effects. The gleaned insights are then integrated into the AI pipeline. Afterwards, we build a deep reinforcement learning framework tailored for drug design, leveraging an autoregressive recurrent neural network model for molecule generation. However, despite the effectiveness of generative model and goal-directed generation with reinforcement learning, a bottleneck lies in the accurate prediction of molecular properties, particularly bioactivity. To address these existing challenges, we conduct an extensive evaluation of representative models using various representations on diverse datasets by training 62,820 models, to unravel key elements underlying molecular property prediction.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798381970326Subjects--Topical Terms:

583819
Pharmacology.
Subjects--Index Terms:

Drug discoveryIndex Terms--Genre/Form:

554714
Electronic books.

An Integrated Artificial Intelligence Pipeline for Drug Discovery.
LDR:03292ntm a22004097 4500 001 1151880
005 20241118135853.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798381970326
035 $a (MiAaPQ)AAI30818526
035 $a AAI30818526
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Deng, Jianyuan. $3 1478719
245 1 3 $a An Integrated Artificial Intelligence Pipeline for Drug Discovery.
264 0 $c 2023
300 $a 1 online resource (200 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
500 $a Advisor: Wang, Fusheng;Chen, Chao;Samaras, Dimitris.
502 $a Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.
504 $a Includes bibliographical references
520 $a Drug discovery is an expensive process in both time and cost with a daunting attrition rate. On average, developing a new drug costs 2.6 billion US dollars and can take more than 10 years. In the past decade, the practice of drug discovery has been undergoing radical transformations in light of the advancement in artificial intelligence (AI) as well as biomedical informatics. In this dissertation, we explore ways to build an integrated AI pipeline for drug discovery.To this end, we first conduct a survey of AI techniques in drug discovery, including model architectures and learning paradigms applicable in molecular property prediction and molecule generation. With the techniques to automate drug design, one key question persists: "what constitutes the desired properties of ideal drug candidates?" In response, we present an informatics-based approach aimed at identifying key pharmacological components underlying drug-drug interactions, which provide insights for drug design. Subsequently, we introduce a large-scale observational study designed to reveal optimal properties underlying opioid analgesics with reduced overdose effects. The gleaned insights are then integrated into the AI pipeline. Afterwards, we build a deep reinforcement learning framework tailored for drug design, leveraging an autoregressive recurrent neural network model for molecule generation. However, despite the effectiveness of generative model and goal-directed generation with reinforcement learning, a bottleneck lies in the accurate prediction of molecular properties, particularly bioactivity. To address these existing challenges, we conduct an extensive evaluation of representative models using various representations on diverse datasets by training 62,820 models, to unravel key elements underlying molecular property prediction.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Pharmacology. $3 583819
650 4 $a Computer science. $3 573171
650 4 $a Bioinformatics. $3 583857
653 $a Drug discovery
653 $a Molecular property prediction
653 $a Model architectures
653 $a Learning paradigms
653 $a Pharmacological components
655 7 $a Electronic books. $2 local $3 554714
690 $a 0800
690 $a 0984
690 $a 0419
690 $a 0715
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a State University of New York at Stony Brook. $b Biomedical Informatics. $3 1478720
773 0 $t Dissertations Abstracts International $g 85-09B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30818526 $z click for full text (PQDT)