서브메뉴
검색
Maximum Entropy Models in Biological Systems
Maximum Entropy Models in Biological Systems
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105256
- ISBN
- 9798265450449
- DDC
- 574.191
- 저자명
- Sarra, Camilla.
- 서명/저자
- Maximum Entropy Models in Biological Systems
- 발행사항
- [Sl] : Princeton University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 108 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-06, Section: B.
- 주기사항
- Advisor: Bialek, William.
- 학위논문주기
- Thesis (Ph.D.)--Princeton University, 2025.
- 초록/해제
- 요약Many living systems show collective behaviors that emerge from interactions among numerous components, operating across scales under noise and constraints. Theoretical biophysics seeks to capture such behaviors by integrating data with physical and statistical models. When modeling the details of the interactions is challenging, a statistical approach may still reveal quantitatively accurate properties. A central tool is the principle of maximum entropy, which provides a systematic way to construct the least-biased probabilistic model consistent with measured constraints.In this dissertation, we apply these concepts to data from the mouse brain, focusing on gene expression and neural activity. For gene expression, we construct the maximum entropy model that matches the average presence of mRNA species and their pairwise correlations. This model accurately predicts higher-order statistics and reveals a probability landscape with multiple local maxima, offering a parameter-free classification criterion for cell types, consistent with existing approaches but grounded in statistical physics.For neural activity, we examine the limitations of a fully connected maximum entropy model in large populations, where under sampling and noise hinder a reliable reconstruction of the full correlation matrix. We show that pruning unreliable constraints yields sparser models that remain predictive, suggesting scalable strategies for modeling increasingly large neural populations.
- 일반주제명
- Biophysics
- 일반주제명
- Physics
- 일반주제명
- Biochemistry
- 일반주제명
- Genetics
- 키워드
- Brain
- 키워드
- Gene expression
- 키워드
- Ising model
- 키워드
- Maximum entropy
- 키워드
- Neural activity
- 기타저자
- Princeton University Physics
- 기본자료저록
- Dissertations Abstracts International. 87-06B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360045
■00520260202105256
■006m o d
■007cr#unu||||||||
■020 ▼a9798265450449
■035 ▼a(MiAaPQ)AAI32278877
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a574.191
■1001 ▼aSarra, Camilla.
■24510▼aMaximum Entropy Models in Biological Systems
■260 ▼a[Sl]▼bPrinceton University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a108 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-06, Section: B.
■500 ▼aAdvisor: Bialek, William.
■5021 ▼aThesis (Ph.D.)--Princeton University, 2025.
■520 ▼aMany living systems show collective behaviors that emerge from interactions among numerous components, operating across scales under noise and constraints. Theoretical biophysics seeks to capture such behaviors by integrating data with physical and statistical models. When modeling the details of the interactions is challenging, a statistical approach may still reveal quantitatively accurate properties. A central tool is the principle of maximum entropy, which provides a systematic way to construct the least-biased probabilistic model consistent with measured constraints.In this dissertation, we apply these concepts to data from the mouse brain, focusing on gene expression and neural activity. For gene expression, we construct the maximum entropy model that matches the average presence of mRNA species and their pairwise correlations. This model accurately predicts higher-order statistics and reveals a probability landscape with multiple local maxima, offering a parameter-free classification criterion for cell types, consistent with existing approaches but grounded in statistical physics.For neural activity, we examine the limitations of a fully connected maximum entropy model in large populations, where under sampling and noise hinder a reliable reconstruction of the full correlation matrix. We show that pruning unreliable constraints yields sparser models that remain predictive, suggesting scalable strategies for modeling increasingly large neural populations.
■590 ▼aSchool code: 0181.
■650 4▼aBiophysics
■650 4▼aPhysics
■650 4▼aBiochemistry
■650 4▼aGenetics
■653 ▼aBrain
■653 ▼aGene expression
■653 ▼aIsing model
■653 ▼aMaximum entropy
■653 ▼aNeural activity
■653 ▼aStatistical mechanics
■690 ▼a0786
■690 ▼a0605
■690 ▼a0487
■690 ▼a0369
■71020▼aPrinceton University▼bPhysics.
■7730 ▼tDissertations Abstracts International▼g87-06B.
■790 ▼a0181
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360045▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
