检索结果-内蒙古大学图书馆

Towards cross-modal causal structure and representation learning 2

学校读者我要写书评

暂无评论

Towards cross-modal causal structure and representation lear...

2nd Machine Learning for Health Symposium, ML4H 2022

作者： Mao, Haiyi Liu, Hongfu Xiaotian Dou, Jason Benos, Panayiotis V. Joint CMU-Pitt PhD Program in Computational Biology University of Pittsburgh United States Michtom School of Computer Science Brandeis University United States Department of Electrical and Computer Engineering University of Pittsburgh United States Department of Epidemiology University of Florida United States

Does the SARS-CoV-2 virus cause patients' chest X-Rays ground-glass opacities? Does an IDH-mutation cause differences in patients' MRI images? Conventional causal discovery algorithms, although well developed to uncover the cause-effect relationships on structured data, cannot elucidate causal relations between unstructured images and structured scalar variables due to the complexity of the former. In this paper, we consider causal discovery between images and structured (scalar) variables. Specifically, we derive low dimensional image representations to analyze with structured variables. We propose a two-module amortized variational algorithm named Cross-Modal Variational Causal representation and structure Learning (CMCL). CMCL jointly learns identifiable representations given a set of independent structured variables and causal relations via formulating latent representations and structured variables into a direct acyclic graph. Moreover, we further enforce counterfactual invariance/variance onto representations. We demonstrate that CMCL outperforms other related methods on synthetic datasets and validate causal relations on semi-synthetic datasets by visualization. © 2022 P.N. Argaw, E. Healey & I.S. Kohane.

关键词： Coronavirus

Random forest model identifies serve strength as a key predictor of tennis match outcome

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Gao, Zijian Kowalczyk, Amanda Darlington School RomeGA United States Carnegie Mellon University University of Pittsburgh PhD Program in Computational Biology PittsburghPA United States

Tennis is a popular sport worldwide, boasting millions of fans and numerous national and international tournaments. Like many sports, tennis has benefitted from the popularity of rigorous record-keeping of game and player information, as well as the growth of machine learning methods for use in sports analytics. Of particular interest to bettors and betting companies alike is potential use of sports records to predict tennis match outcomes prior to match start. We compiled, cleaned, and used the largest database of tennis match information to date to predict match outcome using fairly simple machine learning methods. Using such methods allows for rapid fit and prediction times to readily incorporate new data and make real-time predictions. We were able to predict match outcomes with upwards of 80% accuracy, much greater than predictions using betting odds alone, and identify serve strength as a key predictor of match outcome. By combining prediction accuracies from three models, we were able to nearly recreate a probability distribution based on average betting odds from betting companies, which indicates that betting companies are using similar information to assign odds to matches. These results demonstrate the capability of relatively simple machine learning models to quite accurately predict tennis match outcomes. Copyright © 2019, The Authors. All rights reserved.

关键词： Machine learning

COEM: Cross-Modal Embedding for MetaCell Identification

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Mao, Haiyi Jia, Minxue Dou, Jason Xiaotian Zhang, Haotian Benos, Panayiotis V. Department of Computational and Systems Biology University of Pittsburgh PA United States Joint Carnegie Mellon University University of Pittsburgh PhD Program in Computational Biology United States Department of Electrical and Computer Engineering University of Pittsburgh PA United States

Metacells are disjoint and homogeneous groups of single-cell profiles, representing discrete and highly granular cell states. Existing metacell algorithms tend to use only one modality to infer metacells, even though single-cell multi-omics datasets profile multiple molecular modalities within the same cell. Here, we present Cross-MOdal Embedding for MetaCell Identification (COEM), which utilizes an embedded space leveraging the information of both scATAC-seq and scRNA-seq to perform aggregation, balancing the trade-off between fine resolution and sufficient sequencing coverage. COEM outperforms the state-of-the-art method SEACells by efficiently identifying accurate and well-separated metacells across datasets with continuous and discrete cell types. Furthermore, COEM significantly improves peak-to-gene association analyses, and facilitates complex gene regulatory inference tasks. © 2022, CC BY.

关键词： Embeddings

Dish-trend: Intervention modeling simulator that accounts for trend influences

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Andjelkovic, Stefan Miskov-Zivanov, Natasa Computational and Systems Biology Joint Carnegie Mellon University University of Pittsburgh Computational Biology PhD Program University of Pittsburgh 4420 Bayard St PittsburghPA15213 United States Electrical and Computer Engineering Bioengineering Computational and Systems Biology University of Pittsburgh 4420 Bayard St PittsburghPA15213 United States

关键词： Stochastic systems

Simultaneous identification for synergistic transcription factors and their transcription factor binding sites

学校读者我要写书评

暂无评论

Simultaneous identification for synergistic transcription fa...

5th International Conference on Complex, Intelligent and Software Intensive Systems, CISIS 2011

作者： Tsai, Zing Tsung-Yeh Huang, Grace Tzu-Wei Tsai, Huai-Kuang Institute of Information Science Academia Sinica Taipei 115 Taiwan Bioinformatics Program Taiwan International Graduate Program Taiwan Institute of Biomedical Informatics National Yang-Ming University Taiwan Joint CMU-Pitt PhD Program in Computational Biology University of Pittsburgh Carnegie Mellon University Pittsburgh PA United States Research Center for Information Technology Innovation Taipei 115 Taiwan Computer Science and Engineering National Taiwan Ocean University Keelung 202 Taiwan

ISBN: (纸本)9780769543734

Identifying transcription factor binding sites (TFBSs) is crucial for understanding the mechanism of transcriptional regulation. It is known that transcription factors (TFs) often cooperate to regulate genes. While traditional approaches can be used to discover binding motifs of a group of co-regulated genes, they often fail to accurately assign motifs to the corresponding TFs. Here, we consider two TFs together to infer their TFBSs and their synergistic relationship simultaneously. The basic idea is that if two TFs interact, their TFBSs, if distinct, would be conserved across species and coincided in the promoter regions of the genes they co-regulated. Applying our method to Saccharomyces cerevisiae chromatin immunoprecipitation data, we predicted 110 TF pairs with statistically significant motif assignments. A majority of these TF pairs have literature support to be synergistic, and the designated motifs to TFs match well with their known consensus. We further examined the synergism of predicted TF pairs in seven experimental conditions using ANOVA, and identified significant interactions. © 2011 IEEE.

关键词： Yeast

Mixed graphical models for causal analysis of multi-modal variables

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Sedgewick, Andrew J. Ramsey, Joseph D. Spirtes, Peter Glymour, Clark Benos, Panayiotis V. Department of Computational and Systems Biology University of Pittsburgh School of Medicine PittsburghPA United States Joint Carnegie Mellon University University of Pittsburgh PhD Program in Computational Biology PittsburghPA United States Department of Bioengineering University of Pittsburgh PittsburghPA United States Department of Philosophy Carnegie Mellon University PittsburghPA United States

Graphical causal models are an important tool for knowledge discovery because they can represent both the causal relations between variables and the multivariate probability distributions over the data. Once learned, causal graphs can be used for classification, feature selection and hypothesis generation, while revealing the underlying causal network structure and thus allowing for arbitrary likelihood queries over the data. However, current algorithms for learning sparse directed graphs are generally designed to handle only one type of data (continuous-only or discrete-only), which limits their applicability to a large class of multi-modal biological datasets that include mixed type variables. To address this issue, we developed new methods that modify and combine existing methods for finding undirected graphs with methods for finding directed graphs. These hybrid methods are not only faster, but also perform better than the directed graph estimation methods alone for a variety of parameter settings and data set sizes. Here, we describe a new conditional independence test for learning directed graphs over mixed data types and we compare performances of different graph learning strategies on synthetic data. Copyright © 2017, The Authors. All rights reserved.

关键词： Directed graphs

Intersecting impact of CAG repeat and huntingtin knockout in stem cell-derived cortical neurons

学校读者我要写书评

暂无评论

Neurobiology of Disease 2025年 210卷 106914-106914页

作者： Stocksdale, Jennifer T. Leventhal, Matthew J. Lam, Stephanie Xu, Yu-Xin Wang, Yang Oliver Wang, Keona Q. Thomas, Reuben Faghihmonzavi, Zohreh Raghav, Yogindra Smith, Charlene Wu, Jie Miramontes, Ricardo Sarda, Kanchan Johnston, Heather Shin, Min-Gyoung Huang, Terry Foster, Mikelle Barch, Mariya Amirani, Naufa Paiz, Chris Easter, Lindsay Duderstadt, Erse Vaibhav, Vineet Sundararaman, Niveda Felsenfeld, Dan P. Vogt, Thomas F. Van Eyk, Jennifer Finkbeiner, Steve Kaye, Julia A. Fraenkel, Ernest Thompson, Leslie M. Department of Neurobiology and Behavior UC Irvine Irvine 92677 CA United States MIT PhD Program in Computational and Systems Biology Cambridge 02139 MA United States MIT Department of Biological Engineering Cambridge 02139 MA United States Department of Biological Engineering Massachusetts Institute of Technology Cambridge 02139 MA United States Center for Systems and Therapeutics Gladstone Institutes San Francisco 94158 CA United States Advanced Clinical Biosystems Research Institute Smidt Heart Institute Cedars-Sinai Medical Center Los Angeles 90048 CA United States Institute of Data Science and Biotechnology Gladstone Institutes San Francisco 94158 CA United States Department of Psychiatry and Human Behavior UC Irvine Irvine 92697 CA United States Department of Biological Chemistry UC Irvine Irvine 92697 CA United States Institute for Memory Impairments and Neurological Disorders UC Irvine Irvine 92697 CA United States CHDI Management Inc New York 10001 NY United States Department of Physiology University of California San Francisco San Francisco 94158 CA United States Department of Neurology University of California San Francisco San Francisco 94158 CA United States Taube/Koret Center for Neurodegenerative Disease Research Gladstone Institutes San Francisco 94158 CA United States

Huntington's Disease (HD) is caused by a CAG repeat expansion in the gene encoding huntingtin (HTT). While normal HTT function appears impacted by the mutation, the specific pathways unique to CAG repeat expansion versus loss of normal function are unclear. To understand the impact of the CAG repeat expansion, we evaluated biological signatures of HTT knockout (HTT KO) versus those that occur from the CAG repeat expansion by applying multi-omics, live cell imaging, survival analysis and a novel feature-based pipeline to study cortical neurons (eCNs) derived from an isogenic human embryonic stem cell series (RUES2). HTT KO and the CAG repeat expansion influence developmental trajectories of eCNs, with opposing effects on growth. Network analyses of differentially expressed genes and proteins associated with enriched epigenetic motifs identified subnetworks common to CAG repeat expansion and HTT KO that include neuronal differentiation, cell cycle regulation, and mechanisms related to transcriptional repression, and may represent gain-of-function mechanisms that cannot be explained by HTT loss of function alone. A combination of dominant and loss-of-function mechanisms are likely involved in the aberrant neurodevelopmental and neurodegenerative features of HD that can help inform therapeutic strategies. © 2025

关键词： Embryonic stem cells Features Huntington's disease Machine learning Multi-omics Network analysis Robotic microscopy RUES2

TET proteins safeguard bivalent promoters from de novo methylation in human embryonic stem cells (vol 50, pg 83, 2017)

学校读者我要写书评

暂无评论

NATURE GENETICS 2018年第5期50卷 764-764页

作者： Verma, Nipun Pan, Heng Dore, Louis C. Shukla, Abhijit Li, Qing V. Pelham-Webb, Bobbie Teijeiro, Virginia Gonzalez, Federico Krivtsov, Andrei Chang, Chan-Jung Papapetrou, Eirini P. He, Chuan Elemento, Olivier Huangfu, Danwei Developmental Biology Program Sloan Kettering Institute New York NY USA Weill Graduate School of Medical Sciences at Cornell University/The Rockefeller University/Sloan Kettering Institute Tri-Institutional MD-PhD Program New York NY USA Department of Physiology and Biophysics Englander Institute for Precision Medicine Institute for Computational Biomedicine Weill Cornell Medical College New York NY USA Weill Cornell Graduate School of Medical Sciences Weill Cornell Medical College New York NY USA Department of Chemistry Department of Biochemistry and Molecular Biology Institute for Biophysical Dynamics Howard Hughes Medical Institute University of Chicago Chicago IL USA Louis V. Gerstner Jr. Graduate School of Biomedical Sciences Memorial Sloan Kettering Cancer Center New York NY USA Cancer Biology and Genetics Program Sloan Kettering Institute New York NY USA Department of Oncological Sciences and Black Family Stem Cell Institute Icahn School of Medicine at Mount Sinai New York NY USA

TET enzymes oxidize 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), which can lead to DNA demethylation. However, direct connections between TET-mediated DNA demethylation and transcriptional output are difficult to establish owing to challenges in distinguishing global versus locus-specific effects. Here we show that TET1, TET2 and TET3 triple-knockout (TKO) human embryonic stem cells (hESCs) exhibit prominent bivalent promoter hypermethylation without an overall corresponding decrease in gene expression in the undifferentiated state. Focusing on the bivalent PAX6 locus, we find that increased DNMT3B binding is associated with promoter hypermethylation, which precipitates a neural differentiation defect and failure of PAX6 induction during differentiation. dCas9-mediated locus-specific demethylation and global inactivation of DNMT3B in TKO hESCs partially reverses the hypermethylation at the PAX6 promoter and improves differentiation to neuroectoderm. Taking these findings together with further genome-wide methylation and TET1 and DNMT3B ChIP-seq analyses, we conclude that TET proteins safeguard bivalent promoters from de novo methylation to ensure robust lineage-specific transcription upon differentiation.

关键词： Epigenetics Stem cells

End-to-End Differentiable Molecular Mechanics Force Field Construction

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wang, Yuanqing Fass, Josh Kaminow, Benjamin Herr, John E. Rufa, Dominic Zhang, Ivy Pulido, Iván Henry, Mike Chodera, John D. Computational and Systems Biology Program Sloan Kettering Institute Memorial Sloan Kettering Cancer Center New YorkNY10065 United States Tri-Institutional PhD Program in Computational Biology and Medicine Weill Cornell Medical College Cornell University New YorkNY10065 United States Physiology Biophysics and System Biology Ph.D. Program Weill Cornell Medical College Cornell University New YorkNY10065 United States Program in Creative Writing Division of Humanities and Arts The City College of New York The City University of New York New YorkNY10031 United States

Molecular mechanics (MM) potentials have long been a workhorse of computational chemistry. Leveraging accuracy and speed, these functional forms find use in a wide variety of applications in biomolecular modeling and drug discovery, from rapid virtual screening to detailed free energy calculations. Traditionally, MM potentials have relied on human-curated, inflexible, and poorly extensible discrete chemical perception rules (atom types) for applying parameters to small molecules or biopolymers, making it difficult to optimize both types and parameters to fit quantum chemical or physical property data. Here, we propose an alternative approach that uses graph neural networks to perceive chemical environments, producing continuous atom embeddings from which valence and nonbonded parameters can be predicted using invariance-preserving layers. Since all stages are built from smooth neural functions, the entire process-spanning chemical perception to parameter assignment-is modular and end-to-end differentiable with respect to model parameters, allowing new force fields to be easily constructed, extended, and applied to arbitrary molecules. We show that this approach is not only sufficiently expressive to reproduce legacy atom types, but that it can learn to accurately reproduce and extend existing molecular mechanics force fields. Trained with arbitrary loss functions, it can construct entirely new force fields self-consistently applicable to both biopolymers and small molecules directly from quantum chemical calculations, with superior fidelity than traditional atom or parameter typing schemes. When adapted to simultaneously fit partial charge models, espaloma delivers high-quality partial atomic charges orders of magnitude faster than current best-practices with little inaccuracy. When trained on the same quantum chemical small molecule dataset used to parameterize the Open Force Field ("Parsley") openff-1.2.0 small molecule force field augmented with a peptide dataset

关键词： Atoms