检索结果-内蒙古大学图书馆

HiPoNet: A Topology-Preserving Multi-View Neural Network For High Dimensional Point Cloud and Single-Cell Data

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Viswanath, Siddharth Madhu, Hiren Bhaskar, Dhananjay Kovalic, Jake Johnson, Dave Ying, Rex Tape, Christopher Adelstein, Ian Perlmutter, Michael Krishnaswamy, Smita Department of Computer Science Yale University New Haven United States Department of Genetics Yale University New Haven United States Department of Applied Physics Yale University New Haven United States Program in Computing Boise State University Idaho United States Cell Communication Lab Department of Oncology University College London Cancer Institute London United Kingdom Department of Mathematics Yale University New Haven United States Department of Mathematics Boise State University Idaho United States Computational Biology and Bioinformatics Program Yale University New Haven United States Wu-Tsai Institute Yale University New Haven United States Program for Applied Math Yale University New Haven United States

In this paper, we propose HiPoNet, an end-to-end differentiable neural network for regression, classification, and representation learning on high-dimensional point clouds. Single-cell data can have high dimensionality exceeding the capabilities of existing methods for point clouds which are tailored for 3D data. Moreover, modern single-cell and spatial experiments now yield entire cohorts of datasets (i.e. one on every patient), necessitating models that can process large, high-dimensional point clouds at scale. Most current approaches build a single nearest-neighbor graph, discarding important geometric information. In contrast, HiPoNet forms higher-order simplicial complexes through learnable feature reweighting, generating multiple data views that disentangle distinct biological processes. It then employs simplicial wavelet transforms to extract multiscale features—capturing both local and global topology. We empirically show that these components preserve topological information in the learned representations, and that HiPoNet significantly outperforms state-of-the-art point-cloud and graph-based models on single cell. We also show an application of HiPoNet on spatial transcriptomics datasets using spatial co-ordinates as one of the views. Overall, HiPoNet offers a robust and scalable solution for high-dimensional data analysis. © 2025, CC BY.

关键词： Network topology

Principal Curvatures Estimation with Applications to Single Cell Data

学校读者我要写书评

暂无评论

Principal Curvatures Estimation with Applications to Single ...

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Yanlei Zhang Lydia Mezrag Xingzhi Sun Charles Xu Kincaid Macdonald Dhananjay Bhaskar Smita Krishnaswamy Guy Wolf Bastian Rieck Dept. of Math. & Stat. Université de Montréal Mila – Quebec AI Institute Montréal CA Dept. of Comp. Sci. Yale University Dept. of Genetics Applied Mathematics Program New Haven CT USA Helmholtz Zentrum München Institute of AI for Health Munich BY DE Department of Informatics Université de Fribourg Fribourg CH

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

关键词： Manifolds Sequential analysis Adaptation models Data analysis RNA Transcriptomics Curve fitting Speech processing Tuning Principal component analysis

Principal Curvatures Estimation with Applications to Single Cell Data

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zhang, Yanlei Mezrag, Lydia Sun, Xingzhi Xu, Charles Macdonald, Kincaid Bhaskar, Dhananjay Krishnaswamy, Smita Wolf, Guy Rieck, Bastian Université de Montréal Dept. of Math. & Stat. QC Montréal Canada Mila – Quebec AI Institute MontréalQC Canada Yale University Dept. of Comp. Sci. New HavenCT United States Dept. of Genetics New HavenCT United States Applied Mathematics Program New HavenCT United States Université de Fribourg Department of Informatics Fribourg Switzerland Helmholtz Zentrum München Institute of AI for Health Munich Germany

关键词： Spatio-temporal data

KLF2 maintains lineage fidelity and suppresses CD8 T cell exhaustion during acute LCMV infection

学校读者我要写书评

暂无评论

Science (New York, N.Y.) 2025年第6735期387卷 eadn2337页

作者： Fagerberg, Eric Attanasio, John Dien, Christine Singh, Jaiveer Kessler, Emily A. Abdullah, Leena Shen, Jian Hunt, Brian G. Connolly, Kelli A. De Brouwer, Edward He, Jiaming Iyer, Nivedita R. Buck, Jessica Borr, Emily R. Damo, Martina Foster, Gena G. Giles, Josephine R. Huang, Yina H. Tsang, John S. Krishnaswamy, Smita Cui, Weiguo Joshi, Nikhil S. Department of Immunobiology Yale University School of Medicine New Haven CT United States Program in Computational Biology & Bioinformatics Yale University New Haven CT United States Department of Microbiology and Immunology Geisel School of Medicine at Dartmouth NH United States Department of Pathology Feinberg School of Medicine at Northwestern University Chicago IL United States Department of Genetics and Computer Science Yale University School of Medicine New Haven CT United States Department of Medicine University of Chicago Chicago IL United States Section of Hematology Department of Internal Medicine Yale University School of Medicine New Haven CT United States Department of Systems Pharmacology and Translational Therapeutics University of Pennsylvania Philadelphia PA United States Institute for Immunology Perelman School of Medicine University of Pennsylvania Philadelphia PA United States Parker Institute for Cancer Immunotherapy Perelman School of Medicine University of Pennsylvania Philadelphia PA United States Yale Center for Systems and Engineering Immunology Yale University School of Medicine New Haven CT United States New Haven CT United States Department of Biomedical Engineering Yale University New Haven CT United States Computational Biology and Bioinformatics Program Yale University CT USA New Haven United States Applied Math Program Yale University New Haven CT United States

Naïve CD8 T cells have the potential to differentiate into a spectrum of functional states during an immune response. How these developmental decisions are made and what mechanisms exist to suppress differentiation toward alternative fates remains unclear. We employed in vivo CRISPR-Cas9-based perturbation sequencing to assess the role of ~40 transcription factors (TFs) and epigenetic modulators in T cell fate decisions. Unexpectedly, we found that knockout of the TF Klf2 resulted in aberrant differentiation to exhausted-like CD8 T cells during acute infection. KLF2 was required to suppress the exhaustion-promoting TF TOX and to enable the TF TBET to drive effector differentiation. KLF2 was also necessary to maintain a polyfunctional tumor-specific progenitor state. Thus, KLF2 provides effector CD8 T cell lineage fidelity and suppresses the exhaustion program.

关键词：

Ensemble Estimation of Distributional Functionals via k-Nearest Neighbors

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Moon, Kevin R. Sricharan, Kumar Hero, Alfred O. Genetics Department and Applied Math Program Yale University Xerox PARC EECS Department University of Michigan

The problem of accurate nonparametric estimation of distributional functionals (integral functionals of one or more probability distributions) has received recent interest due to their wide applicability in signal processing, information theory, machine learning, and statistics. In particular, k-nearest neighbor (nn) based methods have received a lot of attention due to their adaptive nature and their relatively low computational complexity. We derive the mean squared error (MSE) convergence rates of leave-one-out k-nn plug-in density estimators of a large class of distributional functionals without boundary correction. We then apply the theory of optimally weighted ensemble estimation to obtain weighted ensemble estimators that achieve the parametric MSE rate under assumptions that are competitive with the state of the art. The asymptotic distributions of these estimators, which are unknown for all other k-nn based distributional functional estimators, are also presented which enables us to perform hypothesis testing. Copyright © 2017, The Authors. All rights reserved.

关键词： Motion compensation

Graph Fourier MMD for Signals on Graphs

学校读者我要写书评

暂无评论

Graph Fourier MMD for Signals on Graphs

2023 International Conference on Sampling Theory and Applications, SampTA 2023

作者： Leone, Samuel Venkat, Aarthi Huguet, Guillaume Tong, Alexander Wolf, Guy Krishnaswamy, Smita Applied Math Program New HavenCT United States Yale University Computational Biology & Bioinformatics Program New HavenCT United States Université de Montréal Mila - Quebec Ai Institute Dept. of Math. & Stat Montreal Canada Université de Montréal Mila-Quebec Ai Institute Dept. of Comp. Sci. & Oper. Res. Montreal Canada Yale University Department of Computer Science Department of Genetics New HavenCT United States

ISBN: (纸本)9798350328851

While numerous methods have been proposed for computing distances between probability distributions in Euclidean space, relatively little attention has been given to computing such distances for distributions on graphs. However, there has been a marked increase in data that either lies on graph (such as protein interaction networks) or can be modeled as a graph (single cell data), particularly in the biomedical sciences. Thus, it becomes important to find ways to compare signals defined on such graphs. Here, we propose Graph Fourier MMD (GFMMD), a novel distance between distributions and signals on graphs. GFMMD is defined via an optimal witness function that is both smooth on the graph and maximizes the difference in expectation between the pair of distributions on the graph. We find an analytical solution to this optimization problem as well as an embedding of distributions that results from this method. We also prove several properties of this method including scale invariance and applicability to disconnected graphs. We showcase it on graph benchmark datasets as well on single cell RNA-sequencing data analysis. In the latter, we use the GFMMD-based gene embeddings to find meaningful gene clusters. We also propose a novel type of score for gene selection called gene localization score which helps select genes for cellular state space characterization. © 2023 IEEE.

关键词： Genes

Graph Fourier MMD for Signals on Graphs

学校读者我要写书评

暂无评论

Graph Fourier MMD for Signals on Graphs

International Conference on Sampling Theory and Applications (SampTA)

作者： Samuel Leone Aarthi Venkat Guillaume Huguet Alexander Tong Guy Wolf Smita Krishnaswamy Applied Math Program New Haven Connecticut Computational Biology & Bioinformatics Program Yale University New Haven Connecticut Dept. of Math. & Stat Mila - Quebec AI Institute Université de Montréal Montreal Canada Dept. of Comp. Sci. & Oper. Res. Mila-Quebec AI Institute Université de Montréal Montreal Canada Department of Computer Science Department of Genetics Yale University New Haven Connecticut

关键词：

Graph Fourier MMD for Signals on Graphs

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Leone, Samuel Venkat, Aarthi Huguet, Guillaume Tong, Alexander Wolf, Guy Krishnaswamy, Smita Applied Math Program New HavenCT United States Computational Biology & Bioinformatics Program Yale University New HavenCT United States Dept. of Math. & Stat Université de Montréal Mila - Quebec AI Institute MontrealQC Canada Dept. of Comp. Sci. & Oper. Res. Université de Montréal Mila - Quebec AI Institute MontrealQC Canada Department of Computer Science Department of Genetics Yale University New HavenCT United States

While numerous methods have been proposed for computing distances between probability distributions in Euclidean space, relatively little attention has been given to computing such distances for distributions on graphs. However, there has been a marked increase in data that either lies on graph (such as protein interaction networks) or can be modeled as a graph (single cell data), particularly in the biomedical sciences. Thus, it becomes important to find ways to compare signals defined on such graphs. Here, we propose Graph Fourier MMD (GFMMD), a novel distance between distributions and signals on graphs. GFMMD is defined via an optimal witness function that is both smooth on the graph and maximizes difference in expectation between the pair of distributions on the graph. We find an analytical solution to this optimization problem as well as an embedding of distributions that results from this method. We also prove several properties of this method including scale invariance and applicability to disconnected graphs. We showcase it on graph benchmark datasets as well on single cell RNA-sequencing data analysis. In the latter, we use the GFMMD-based gene embeddings to find meaningful gene clusters. We also propose a novel type of score for gene selection called gene localization score which helps select genes for cellular state space characterization. Copyright © 2023, The Authors. All rights reserved.

关键词： Genes

Neural FIM for learning Fisher information metrics from point cloud data

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Fasina, Oluwadamilola Huguet, Guillaume Tong, Alexander Zhang, Yanlei Wolf, Guy Nickel, Maximilian Adelstein, Ian Krishnaswamy, Smita Applied Math Program Yale University New HavenCT United States Mila - Quebec AI Institute MontrealQC Canada Department of Mathematics and Statistics Université de Montréal MontrealQC Canada Department of Computer Science and Operations Research Université de Montréal MontrealQC Canada FAIR Meta AI United States Department of Math Yale University New HavenCT United States Department of Computer Science Yale University New HavenCT United States Department of Genetics Yale University New HavenCT United States

Although data diffusion embeddings are ubiquitous in unsupervised learning and have proven to be a viable technique for uncovering the underlying intrinsic geometry of data, diffusion embeddings are inherently limited due to their discrete nature. To this end, we propose neural FIM, a method for computing the Fisher information metric (FIM) from point cloud data - allowing for a continuous manifold model for the data. Neural FIM creates an extensible metric space from discrete point cloud data such that information from the metric can inform us of manifold characteristics such as volume and geodesics. We demonstrate Neural FIM's utility in selecting parameters for the PHATE visualization method as well as its ability to obtain information pertaining to local volume illuminating branching points and cluster centers embeddings of a toy dataset and two single-cell datasets of IPSC reprogramming and PBMCs (immune cells). Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings