國立虎尾科技大學 |

Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity./
Author:	Zatorski, Nicole.
Description:	1 online resource (216 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-01, Section: B.
Contained By:	Dissertations Abstracts International85-01B.
Subject:	Computational chemistry. -
Online resource:	click for full text (PQDT)
ISBN:	9798379796013

Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity.
Zatorski, Nicole.

Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity. - 1 online resource (216 pages)

Source: Dissertations Abstracts International, Volume: 85-01, Section: B.

Thesis (Ph.D.)--Icahn School of Medicine at Mount Sinai, 2023.

Includes bibliographical references

Drugs with unexpected and severe toxicities cause immense risk to patients. Many strategies, both experimental and computational, have been devised to anticipate and understand the undesirable effects of compounds on the body. Even with these tools however, many drugs still exhibit intolerable safety profiles in human testing and long-term use within the population. It is clear that additional strategies are still needed to accurately predict drug toxicity. A promising approach to addressing the challenge of drug toxicity prediction, which this thesis illustrates, incorporates protein structure-based data with machine learning techniques. Structural features have previously demonstrated their utility for revealing patterns in many biological phenomena such as protein-protein interactions and genetic disease protein networks. This thesis presents a tool that determines structural features and, in combination with machine learning approaches, predicts trends in human tissue identity, cancer classification, as well as drug toxicity.This thesis describes a standardized method for generating protein structural features in part I and applies this tool, along with feature selection and ensemble learning, to address the challenge of drug toxicity prediction in part II. Due to the myriad levels of severity and types of toxicity, labeling a drug with a toxicity status is a non-trivial task. Therefore, this thesis incorporates the toxicity evaluation of drugs as determined by regulatory bodies of experts such as the FDA. Drug withdrawal from the market is thus treated an indication of severe or intolerable drug toxicity. Characteristics of compounds with severe toxicity are compared to those of drugs, which are still available on the market using a variety of feature selection techniques. These approaches look for variability in underlying features between the two groups and determine ranges of expected values for these characteristics. The insights encompassed in this work generate guidelines for drug toxicity assessment during the drug development process. Further advances are made to the field of toxicity prediction with the development of a drug toxicity classification ensemble algorithm trained on drug target features as well as drug chemical features. This tool predicts the likelihood of withdrawal from the market of compounds currently under investigation in clinical trials. The structural feature generation tool, the web server, and the machine learning based characterization of toxic drugs have a variety of applications, particularly in drug development.

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

Mode of access: World Wide Web

ISBN: 9798379796013Subjects--Topical Terms:

1190332
Computational chemistry.
Subjects--Index Terms:

Clinical trialsIndex Terms--Genre/Form:

554714
Electronic books.

Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity.
LDR:04019ntm a22003977 4500 001 1149347
005 20241015112522.5
006 m o d
007 cr bn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798379796013
035 $a (MiAaPQ)AAI30568087
035 $a AAI30568087
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Zatorski, Nicole. $3 1475541
245 1 0 $a Integrating Machine Learning and Protein Structural Features to Predict Drug Toxicity.
264 0 $c 2023
300 $a 1 online resource (216 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: 85-01, Section: B.
500 $a Advisor: Schlessinger, Avner;Sobie, Eric.
502 $a Thesis (Ph.D.)--Icahn School of Medicine at Mount Sinai, 2023.
504 $a Includes bibliographical references
520 $a Drugs with unexpected and severe toxicities cause immense risk to patients. Many strategies, both experimental and computational, have been devised to anticipate and understand the undesirable effects of compounds on the body. Even with these tools however, many drugs still exhibit intolerable safety profiles in human testing and long-term use within the population. It is clear that additional strategies are still needed to accurately predict drug toxicity. A promising approach to addressing the challenge of drug toxicity prediction, which this thesis illustrates, incorporates protein structure-based data with machine learning techniques. Structural features have previously demonstrated their utility for revealing patterns in many biological phenomena such as protein-protein interactions and genetic disease protein networks. This thesis presents a tool that determines structural features and, in combination with machine learning approaches, predicts trends in human tissue identity, cancer classification, as well as drug toxicity.This thesis describes a standardized method for generating protein structural features in part I and applies this tool, along with feature selection and ensemble learning, to address the challenge of drug toxicity prediction in part II. Due to the myriad levels of severity and types of toxicity, labeling a drug with a toxicity status is a non-trivial task. Therefore, this thesis incorporates the toxicity evaluation of drugs as determined by regulatory bodies of experts such as the FDA. Drug withdrawal from the market is thus treated an indication of severe or intolerable drug toxicity. Characteristics of compounds with severe toxicity are compared to those of drugs, which are still available on the market using a variety of feature selection techniques. These approaches look for variability in underlying features between the two groups and determine ranges of expected values for these characteristics. The insights encompassed in this work generate guidelines for drug toxicity assessment during the drug development process. Further advances are made to the field of toxicity prediction with the development of a drug toxicity classification ensemble algorithm trained on drug target features as well as drug chemical features. This tool predicts the likelihood of withdrawal from the market of compounds currently under investigation in clinical trials. The structural feature generation tool, the web server, and the machine learning based characterization of toxic drugs have a variety of applications, particularly in drug development.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computational chemistry. $3 1190332
650 4 $a Pharmacology. $3 583819
650 4 $a Bioinformatics. $3 583857
653 $a Clinical trials
653 $a Drug toxicity
653 $a Machine learning
653 $a Rational drug design
653 $a Structural features
655 7 $a Electronic books. $2 local $3 554714
690 $a 0219
690 $a 0419
690 $a 0715
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Icahn School of Medicine at Mount Sinai. $b Pharmacology and System Biology. $3 1474870
773 0 $t Dissertations Abstracts International $g 85-01B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30568087 $z click for full text (PQDT)