arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2308.08060stat.MLcs.LGmath.AGq-bio.GNstat.AP

基于零膨胀泊松模型与共识聚合的鲁棒贝叶斯张量分解

Robust Bayesian Tensor Factorization with Zero-Inflated Poisson Model and Consensus Aggregation

Broad Institute of MIT and Harvard（麻省理工学院与哈佛大学博德研究所）；Klarman Cell Observatory, Broad Institute of MIT and Harvard（麻省理工学院与哈佛大学博德研究所克拉曼细胞观测站）；Brigham and Women’s Hospital（布莱根妇女医院）；Harvard Medical School（哈佛医学院）；University of Massachusetts Boston（马萨诸塞大学波士顿分校）

Daniel Chafamo, Vignesh Shanmugam, Neriman Tokcan

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对零膨胀计数数据及张量分解结果不稳定的问题，提出ZIPTF与共识聚合C-ZIPTF，在合成和真实scRNA-seq数据上提升了重建精度、一致性及生物程序恢复能力。

摘要：

张量分解（TF）是高效表示和分析多维数据的强大工具。然而，基于最大似然估计的经典TF方法在应用于零膨胀计数数据（例如单细胞RNA测序（scRNA-seq）数据）时表现不佳。此外，TF固有的随机性会导致因子在重复运行之间发生变化，从而使结果的解释和可复现性具有挑战性。本文提出零膨胀泊松张量分解（Zero Inflated Poisson Tensor Factorization，ZIPTF），这是一种用于分解具有过量零的高维计数数据的新方法。为解决随机性挑战，我们提出共识零膨胀泊松张量分解（Consensus Zero Inflated Poisson Tensor Factorization，C-ZIPTF），它将ZIPTF与基于共识的元分析相结合。我们在合成零膨胀计数数据以及合成和真实scRNA-seq数据上评估了所提出的ZIPTF和C-ZIPTF。就零膨胀数据的重建精度而言，ZIPTF始终优于基线矩阵和张量分解方法。当过量零的概率较高时，ZIPTF的精度最高可提升2.4倍。此外，C-ZIPTF显著提高了分解的一致性和准确性。在合成和真实scRNA-seq数据上测试时，ZIPTF和C-ZIPTF始终能够恢复已知且具有生物学意义的基因表达程序。

英文摘要：

Tensor factorizations (TF) are powerful tools for the efficient representation and analysis of multidimensional data. However, classic TF methods based on maximum likelihood estimation underperform when applied to zero-inflated count data, such as single-cell RNA sequencing (scRNA-seq) data. Additionally, the stochasticity inherent in TFs results in factors that vary across repeated runs, making interpretation and reproducibility of the results challenging. In this paper, we introduce Zero Inflated Poisson Tensor Factorization (ZIPTF), a novel approach for the factorization of high-dimensional count data with excess zeros. To address the challenge of stochasticity, we introduce Consensus Zero Inflated Poisson Tensor Factorization (C-ZIPTF), which combines ZIPTF with a consensus-based meta-analysis. We evaluate our proposed ZIPTF and C-ZIPTF on synthetic zero-inflated count data and synthetic and real scRNA-seq data. ZIPTF consistently outperforms baseline matrix and tensor factorization methods in terms of reconstruction accuracy for zero-inflated data. When the probability of excess zeros is high, ZIPTF achieves up to $2.4\times$ better accuracy. Additionally, C-ZIPTF significantly improves the consistency and accuracy of the factorization. When tested on both synthetic and real scRNA-seq data, ZIPTF and C-ZIPTF consistently recover known and biologically meaningful gene expression programs.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑