部分特征拓指导可靠的无scRNA集成
Chuan He1, Paraskevas Filippidis2, Steven H Kleinstein2,3,4
1Department of Biostatistics, Yale School of Public Health, New Haven, CT, US.
我们介绍了scCRAFT,这是一个用于单细胞RNA测序 (scRNA-seq) 批量集成的新型自编码器. 这种方法有效地纠正批量效应,同时保持重要的生物异质性,以准确进行细胞类型分析.
科学领域:
- 计算生物学 计算生物学
- 基因组学就是基因组学.
- 生物信息学是一种生物信息学.
背景情况:
- 单细胞RNA测序 (scRNA-seq) 提供了高分辨率的生物学见解.
- scRNA-seq批量集成对于下游分析至关重要,但具有挑战性.
- 现有的方法努力纠正批量效应,同时保持细胞类型异质性.
研究的目的:
- 开发一种可靠的方法,用于scRNA-seq数据中的多批整合.
- 为了应对可靠批量校正的挑战,而不会失去生物信号.
- 在集成的scRNA-seq数据集中提高细胞类型识别的准确性.
主要方法:
- 提出了scCRAFT,这是一个用于scRNA-seq批量集成的自编码模型.
- 集成了三个损失组件:重建,多域适应和双分辨率三重损失.
- 双分辨率三重损失可以抵消异质细胞分布中的过度校正.
主要成果:
- 在模拟中,scCRAFT有效处理不平衡的批量和罕见的细胞类型.
- 该模型针对批量特定的细胞表型.
- scCRAFT在各种现实世界 scRNA-seq 数据集上的性能优于最先进的方法.
结论:
- scCRAFT为scRNA-seq数据提供可靠的批次校正.
- 该方法成功地保持了对细胞类型定义至关重要的生物异质性.
- scCRAFT推进了用于scRNA-seq分析的多批次集成.
更多相关视频
06:58Global Identification of Co-Translational Interaction Networks by Selective Ribosome Profiling
Published on: October 7, 2021
09:04Sequence-specific and Selective Recognition of Double-stranded RNAs over Single-stranded RNAs by Chemically Modified Peptide Nucleic Acids
Published on: September 21, 2017
相关概念视频
Tail-anchoring of Proteins in the ER Membrane
Directing Proteins to the Rough Endoplasmic Reticulum
Conserved Binding Sites
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
Insertion of Single-pass Transmembrane Proteins in the RER
Integral transmembrane proteins possess transmembrane and extra membrane domains. The transmembrane domains are primarily made of 20-25 hydrophobic amino acids arranged in a helical secondary confirmation. These...
Nuclear Localization Signals and Import
RNA Splicing
