小柯机器人

近日，德国慕尼黑工业大学Fabian J. Theis开发出通过转移学习将单细胞数据映射到参考图集的新方法。这一研究成果于2021年8月30日在线发表在国际学术期刊《自然—生物技术》上。

研究人员报道了一种深度学习策略，用于在称为单细胞结构手术（scArches）的参考之上映射查询数据集。scArches使用转移学习和参数优化来实现高效、分散、迭代的参考构建和新数据集，而无需分享原始数据。利用小鼠大脑、胰腺、免疫和全生物体图谱的例子，研究人员表明scArches保留了生物状态信息，同时消除了批次效应，尽管使用的参数比从头整合少四个数量级。

最后，scArches在映射到健康参照物时保留了COVID-19的疾病变异，从而能够发现疾病特定的细胞状态。scArches将通过实现参照物图谱的迭代构建、更新、共享和有效使用来促进合作项目。

据了解，大型单细胞图谱作为分析小规模研究的参考。然而，由于数据集之间的批次效应、计算资源的有限可用性和原始数据的共享限制，从参考数据中学习是复杂的。

附：英文原文

Title: Mapping single-cell data to reference atlases by transfer learning

Author: Lotfollahi, Mohammad, Naghipourfar, Mohsen, Luecken, Malte D., Khajavi, Matin, Bttner, Maren, Wagenstetter, Marco, Avsec, iga, Gayoso, Adam, Yosef, Nir, Interlandi, Marta, Rybakov, Sergei, Misharin, Alexander V., Theis, Fabian J.

Issue&Volume: 2021-08-30

Abstract: Large single-cell atlases are now routinely generated to serve as references for analysis of smaller-scale studies. Yet learning from reference data is complicated by batch effects between datasets, limited availability of computational resources and sharing restrictions on raw data. Here we introduce a deep learning strategy for mapping query datasets on top of a reference called single-cell architectural surgery (scArches). scArches uses transfer learning and parameter optimization to enable efficient, decentralized, iterative reference building and contextualization of new datasets with existing references without sharing raw data. Using examples from mouse brain, pancreas, immune and whole-organism atlases, we show that scArches preserves biological state information while removing batch effects, despite using four orders of magnitude fewer parameters than de novo integration. scArches generalizes to multimodal reference mapping, allowing imputation of missing modalities. Finally, scArches retains coronavirus disease 2019 (COVID-19) disease variation when mapping to a healthy reference, enabling the discovery of disease-specific cell states. scArches will facilitate collaborative projects by enabling iterative construction, updating, sharing and efficient use of reference atlases. 

DOI: 10.1038/s41587-021-01001-7

Source: https://www.nature.com/articles/s41587-021-01001-7

Nature Biotechnology：《自然—生物技术》，创刊于1996年。隶属于施普林格·自然出版集团，最新IF：68.164
官方网址：https://www.nature.com/nbt/
投稿链接：https://mts-nbt.nature.com/cgi-bin/main.plex