서브메뉴
검색
Mapping and Dissecting Protein Sequence-Function Landscapes With Microfluidics and Deep-Learning
Mapping and Dissecting Protein Sequence-Function Landscapes With Microfluidics and Deep-Learning
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260311091513.5
- ISBN
- 9798280752917
- DDC
- 572.6
- 서명/저자
- Mapping and Dissecting Protein Sequence-Function Landscapes With Microfluidics and Deep-Learning / Duncan Fraser Muir
- 발행사항
- [Sl] : University of California, San Francisco, 2025
- 형태사항
- 1 electronic resource (213 pages)
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-12, Section: B.
- 주기사항
- Advisors: Capra, John A. Committee members: Pinney, Margaux; Fraser, James.
- 학위논문주기
- - Ph.D. : University of California, San Francisco, 2025.
- 초록/해제
- 요약The quantitative relationship between protein sequence and protein function is often represented as a landscape. Protein sequence-function landscapes have many important applications, from engineering carbon-fixing enzymes in bioremediation to combating drug resistance. The vast expanse of protein sequence space renders it impossible to explore entirely. In the age of rapid development of high-throughput technologies and machine learning algorithms, a synergistic combination of experimentation and computation is poised to strike out on a broad expedition to map a sequence-catalysis landscape across enzyme evolution. In this thesis, I build on advances in microfluidic biochemistry and machine learning to map and dissect the sequence-catalysis landscape of a model enzyme, Adenylate Kinase (ADK). In Chapter 1, I will provide a conceptual framework for sequence-function landscapes, discuss their application, and give an overview of exemplary high-throughput technologies and machine learning techniques used to model sequence-function relationships. In Chapter 2, through Evolutionary-Scale Enzymology, we extend the High-Throughput Microfluidic Enzyme Kinetics (HT-MEK) platform to measure catalytic parameters for hundreds of orthologous and mutant ADKs. We then dissect the topology, navigability, and mechanistic underpinnings of the landscape. We then demonstrate the utility of our dataset in improving performance in ADK catalytic turnover prediction compared to a larger model trained on collated literature data for all enzymes. Chapter 3 explores the application of Protein Language Models (PLMs) in greater depth, evaluating the utility of their learned representations for supervised modeling of ADK catalytic turnover. We examine multiple PLMs, embedding pooling strategies, and fine-tuning techniques. Lastly, in Chapter 4, I will close with outstanding questions in the field that can be addressed via further explorations of protein sequence space.
- 언어주기
- English
- 일반주제명
- Biochemistry
- 일반주제명
- Computer science
- 일반주제명
- Pharmacology
- 키워드
- Enzymology
- 키워드
- Machine-learning
- 키워드
- Microfluidics
- 키워드
- Drug resistance
- 키워드
- Adenylate Kinase
- 기타저자
- University of California, San Francisco Biophysics
- 기본자료저록
- Dissertations Abstracts International. 86-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260311s2025 us eng d■001000017357111
■00520260311091513.5
■006m o d
■007cr|nu||||||||
■020 ▼a9798280752917
■040 ▼aMiAaPQD▼beng▼cMiAaPQD▼erda
■082 ▼a572.6
■1001 ▼aMuir, Duncan Fraser▼eauthor.▼0(orcid)0000-0003-0170-5937
■24510▼aMapping and Dissecting Protein Sequence-Function Landscapes With Microfluidics and Deep-Learning ▼cDuncan Fraser Muir
■260 ▼a[Sl]▼bUniversity of California, San Francisco▼c2025
■264 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a1 electronic resource (213 pages)
■336 ▼atext▼btxt▼2rdacontent
■337 ▼acomputer▼bc▼2rdamedia
■338 ▼aonline resource▼bcr▼2rdacarrier
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-12, Section: B.
■500 ▼aAdvisors: Capra, John A. Committee members: Pinney, Margaux; Fraser, James.
■5021 ▼bPh.D.▼cUniversity of California, San Francisco▼d2025.
■520 ▼aThe quantitative relationship between protein sequence and protein function is often represented as a landscape. Protein sequence-function landscapes have many important applications, from engineering carbon-fixing enzymes in bioremediation to combating drug resistance. The vast expanse of protein sequence space renders it impossible to explore entirely. In the age of rapid development of high-throughput technologies and machine learning algorithms, a synergistic combination of experimentation and computation is poised to strike out on a broad expedition to map a sequence-catalysis landscape across enzyme evolution. In this thesis, I build on advances in microfluidic biochemistry and machine learning to map and dissect the sequence-catalysis landscape of a model enzyme, Adenylate Kinase (ADK). In Chapter 1, I will provide a conceptual framework for sequence-function landscapes, discuss their application, and give an overview of exemplary high-throughput technologies and machine learning techniques used to model sequence-function relationships. In Chapter 2, through Evolutionary-Scale Enzymology, we extend the High-Throughput Microfluidic Enzyme Kinetics (HT-MEK) platform to measure catalytic parameters for hundreds of orthologous and mutant ADKs. We then dissect the topology, navigability, and mechanistic underpinnings of the landscape. We then demonstrate the utility of our dataset in improving performance in ADK catalytic turnover prediction compared to a larger model trained on collated literature data for all enzymes. Chapter 3 explores the application of Protein Language Models (PLMs) in greater depth, evaluating the utility of their learned representations for supervised modeling of ADK catalytic turnover. We examine multiple PLMs, embedding pooling strategies, and fine-tuning techniques. Lastly, in Chapter 4, I will close with outstanding questions in the field that can be addressed via further explorations of protein sequence space.
■546 ▼aEnglish
■590 ▼aSchool code: 0034
■650 4▼aBiochemistry
■650 4▼aComputer science
■650 4▼aPharmacology
■653 ▼aEnzymology
■653 ▼aMachine-learning
■653 ▼aMicrofluidics
■653 ▼aDrug resistance
■653 ▼aAdenylate Kinase
■7102 ▼aUniversity of California, San Francisco▼bBiophysics.▼edegree granting institution.
■7201 ▼aCapra, John A.▼edegree supervisor.
■7730 ▼tDissertations Abstracts International▼g86-12B.
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357111▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
