scSemiGCN:从具有极限监督的耐噪图形神经网络中增强细胞类型注释
Jue Yang1, Weiwen Wang2, Xiwen Zhang3
1School of Mathematics, Sun Yat-sen University, Guangzhou 510000, China.
Bioinformatics (Oxford, England)
|February 17, 2024
概括
本研究介绍了scSemiGCN,这是一种新的图形卷积网络方法,用于在单细胞RNA测序数据中强大的细胞类型注释. 它有效地解决了噪声和批量效应,改善了细胞异质性的发现,即使是有限的标记细胞.
科学领域:
- 基因组学就是基因组学.
- 生物信息学是一种生物信息学.
- 计算生物学 计算生物学
背景情况:
- 细胞类型的注释对于理解单细胞RNA测序 (scRNA-seq) 数据中的细胞异质性至关重要.
- scRNA-seq数据经常受到低信号噪声比率,批量效应和丢失的影响,阻碍了准确的细胞类型识别.
- 现有的无监督和半监督方法在发现强大的细胞类型模式方面面临挑战.
研究的目的:
- 为scRNA-seq数据开发一种强大而高效的半监督细胞类型注释方法.
- 为了克服噪音数据,批量效应和有限的标记样本所带来的局限性.
- 通过改进细胞类型注释,增强细胞异质性的发现.
主要方法:
- 提出了scSemiGCN,一种基于图形卷积网络的方法,用于单元类型的注释.
- 采用了无线网络结构来识别可靠的细胞对细胞连接.
- 对于没有注释的单元格,利用伪标签生成,然后进行监督对比学习以提炼数据.
- 执行消息传递在半监督注释的denoised网络中的精细特征.
主要成果:
- scSemiGCN在多个数据集中表现出细胞类型注释的有效性和效率.
- 该方法与其他六种方法相比,表现优越,特别是在极其有限的监督下.
- 验证了准确识别细胞类型的能力,尽管数据噪声和批量效应.
结论:
- scSemiGCN提供了一个强大的解决方案,用于在具有挑战性的scRNA-seq数据集中进行半监督的细胞类型注释.
- 该方法有效处理数据噪声和批量效应,提高细胞异质性分析的准确性.
- scSemiGCN为研究人员分析单细胞数据提供了一个有价值的工具.
相关概念视频
Neural Regulation
45.0K
Digestion begins with a cephalic phase that prepares the digestive system to receive food. When our brain processes visual or olfactory information about food, it triggers impulses in the cranial nerves innervating the salivary glands and stomach to prepare for food.
45.0K
Neural Circuits
3.3K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
3.3K


