在单细胞转录组测序中一个重要的方面就是定义聚类得来的cluster，一个思考的方向就是根据marker基因来定义cluster的细胞类型（cell type）。怎样才能建立cluster和细胞类型之间的关系呢？根据前人研究的结果，我们已经知道许多细胞类型的marker基因，如果这些marker基因在某一个cluster中高表达或者又是它的one-to-others的差异基因，那么是不是就可以定义这个cluster为marker基因对应的细胞类型了呢？我们说是可以的。
下面是几个已知的单细胞marker基因数据库，点相应的加粗字体可以直达数据库官网：
Mouse Cell Atlas
CellMarker
BD||cd_marker_handbook
CancerSEA
PanglaoDB
haemosphere
造血过程关键基因表达谱数据库。
Haemosphere is a data portal designed for the bench biologist, providing intuitive interfaces to access and analyse the Haemopedia datasets and some other key datasets of interest to the haematopoiesis community.
The user can search for genes of interest, view their expression profiles across a range of cell types within a dataset, perform differential expression analyses and manage gene sets easily. The main principle behind Haemosphere’s design is to provide an exploratory tool, with minimum filtering and interpretation, while making the presentation of this data very accessible.
