Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets
Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets
复制标题
DOI：
10.1038/s41467-019-13055-y
复制
发表时间：
2019-11-28
影响因子：
16.6
通讯作者：
Snyder-Cappione, Jennifer E.
中科院分区：
文献类型：
--
作者：
Belkina, Anna C.;Ciccolella, Christopher O.;Snyder-Cappione, Jennifer E.
从高维单细胞数据集中准确、全面地提取信息需要忠实的可视化来评估生物群体。最先进的非线性降维算法 t-SNE 需要多种启发式方法，并且在投影数百万个单元时无法生成数据集的清晰表示。我们开发了 opt-SNE，这是一个用于 t-SNE 参数选择的自动化工具包，它利用实时 Kullback-Leibler 散度评估，以特定于数据集的方式定制梯度下降迭代的早期夸大和总数。早期夸大的精确校准以及梯度下降学习率的 opt-SNE 调整极大地缩短了计算时间，并实现了大型细胞计数和转录组数据集的高质量可视化，克服了具有硬编码参数的分析工具的局限性，这些参数通常会产生分辨率不佳或误导性的荧光和质谱细胞计数数据图。总之，opt-SNE 可在 t-SNE 空间中实现卓越的数据分辨率，从而实现更准确的数据解释。
Accurate and comprehensive extraction of information from high-dimensional single cell datasets necessitates faithful visualizations to assess biological populations. A state-of-the-art algorithm for non-linear dimension reduction, t-SNE, requires multiple heuristics and fails to produce clear representations of datasets when millions of cells are projected. We develop opt-SNE, an automated toolkit for t-SNE parameter selection that utilizes Kullback-Leibler divergence evaluation in real time to tailor the early exaggeration and overall number of gradient descent iterations in a dataset-specific manner. The precise calibration of early exaggeration together with opt-SNE adjustment of gradient descent learning rate dramatically improves computation time and enables high-quality visualization of large cytometry and transcriptomics datasets, overcoming limitations of analysis tools with hard-coded parameters that often produce poorly resolved or misleading maps of fluorescent and mass cytometry data. In summary, opt-SNE enables superior data resolution in t-SNE space and thereby more accurate data interpretation.