Model-based autoencoders for imputing discrete single-cell RNA-seq data.

Model-based autoencoders for imputing discrete single-cell RNA-seq data.
复制标题

DOI:
10.1016/j.ymeth.2020.09.010
复制
发表时间:
2021-08
期刊:
Methods (San Diego, Calif.)
影响因子:
--
通讯作者:
Wei Z
Wei Z
中科院分区:
其他
文献类型:
--
作者:
Tian T;Min MR;Wei Z

文献摘要

参考文献

相似文献

深度神经网络已被广泛应用于缺失数据填补。然而,大多数现有的研究都集中在连续数据的插补,而离散数据的插补是探索不足。离散数据在真实的世界中是很常见的,尤其是在生物信息学、遗传学和生物化学等研究领域。特别是,大量最近的基因组数据是由单细胞RNA测序(scRNA-seq)技术产生的离散计数数据。大多数scRNA-seq研究产生离散矩阵,其具有普遍的“假”零计数观察值(缺失值)。为了使下游分析更有效,通常将恢复缺失值的插补作为预处理scRNA-seq数据的第一步。在本文中,我们提出了一种新的基于零膨胀负二项(ZINB)模型的自动编码器,用于输入离散scRNA-seq数据。我们的方法的新颖性是双重的。首先,除了优化ZINB可能性之外,我们还建议通过使用Gumbel-Softmax分布来明确地对导致缺失值的辍学事件进行建模。其次,相对于原始计数矩阵进一步优化零膨胀重建。仿真数据集上的大量实验表明,零膨胀重建显着提高插补精度。真实的数据实验表明,该方法能有效区分不同细胞类型,提高差异表达分析的准确性。
Deep neural networks have been widely applied for missing data imputation. However, most existing studies have been focused on imputing continuous data, while discrete data imputation is under-explored. Discrete data is common in real world, especially in research areas of bioinformatics, genetics, and biochemistry. In particular, large amounts of recent genomic data are discrete count data generated from single-cell RNA sequencing (scRNA-seq) technology. Most scRNA-seq studies produce a discrete matrix with prevailing ‘false’ zero count observations (missing values). To make downstream analyses more effective, imputation, which recovers the missing values, is often conducted as the first step in pre-processing scRNA-seq data. In this paper, we propose a novel Zero-Inflated Negative Binomial (ZINB) model-based autoencoder for imputing discrete scRNA-seq data. The novelties of our method are twofold. First, in addition to optimizing the ZINB likelihood, we propose to explicitly model the dropout events that cause missing values by using the Gumbel-Softmax distribution. Second, the zero-inflated reconstruction is further optimized with respect to the raw count matrix. Extensive experiments on simulation datasets demonstrate that the zero-inflated reconstruction significantly improves imputation accuracy. Real data experiments show that the proposed imputation can enhance separating different cell types and improve the accuracy of differential expression analysis.
DOI: 10.1016/j.cell.2018.05.061
发表时间: 2018-07-26
期刊: Cell
影响因子: 64.5
作者:
van Dijk D;Sharma R;Nainys J;Yim K;Kathail P;Carr AJ;Burdziak C;Moon KR;Chaffer CL;Pattabiraman D;Bierie B;Mazutis L;Wolf G;Krishnaswamy S;Pe'er D
通讯作者: Pe'er D
从单细胞RNA-seq数据中提取信号的一般而灵活的方法。
DOI: 10.1038/s41467-017-02554-5
发表时间: 2018-01-18
影响因子: 16.6
作者:
Risso D;Perraudeau F;Gribkova S;Dudoit S;Vert JP
通讯作者: Vert JP
DOI: 10.1038/nmeth.2967
发表时间: 2014-07
期刊: NATURE METHODS
影响因子: 48
作者:
Kharchenko, Peter V.;Silberstein, Lev;Scadden, David T.
通讯作者: Scadden, David T.
DOI: 10.1038/s41592-018-0033-z
发表时间: 2018-07
期刊: Nature methods
影响因子: 48
作者:
Huang M;Wang J;Torre E;Dueck H;Shaffer S;Bonasio R;Murray JI;Raj A;Li M;Zhang NR
通讯作者: Zhang NR
使用纳米液液滴对单个细胞的高度平行基因组表达分析。
DOI: 10.1016/j.cell.2015.05.002
发表时间: 2015-05-21
期刊: Cell
影响因子: 64.5
作者:
Macosko EZ;Basu A;Satija R;Nemesh J;Shekhar K;Goldman M;Tirosh I;Bialas AR;Kamitaki N;Martersteck EM;Trombetta JJ;Weitz DA;Sanes JR;Shalek AK;Regev A;McCarroll SA
通讯作者: McCarroll SA