A multiome analysis of sn-RNA and sn-ATAC of hypothalamus cells.
Two hypotlamus samples from mouse [1] were used as test data, each contaning both RNA and ATAC.
The main goal is to identify hypothalamus celltypes and investigate sub-population. The bioinformatics goal is to adapt Seurat [2] piepeline to be used in samples of hypothalamus.
Computational performance of two ways for merging the samples were analysed, the first aggregating sample with cellranger aggr. The second using Serat function merge.
A reference RNA expression map was used (top left figure) to identify the celltypes. Anchor cells were used to annotate ATAC map and to transfer annotation to multimodal RNA+ATAC map.
All celltypes were considered for analysis as well as the corresponding markers identifying each. Noteworthy, Pdgfra is a good marker for OPCs in this testdata.
We next investigated sub-popualations of neuronal cells. In the paper describind the data [1], authors described 33 neuronal su-bpopulations using 6 mouse samples and 100% of the filtered cells. In our test case, ~17 clusters are seen.
Neuronal cells were sublected for clusterring and further investigation.
Neuronal sub-poppulations were investigated for magnocellular population using markers Oxt, Avp and Caprin2. Interesting, clusters 9 and 10 show both, differential ATAC peaks and differential RNA expr.
Noteworthy, clusters 9 and 10 can be considered for further investigation for magnocellular populations.
[1] Integrative single-cell characterization of hypothalamus sex-differential and obesity-associated genes and regulatory elements HP Nguyen, CSY Chan, DL Cintron, R Sheng, L Harshman, M Nobuhara, A Ushiki, C Biellak… bioRxiv, 2022•biorxiv.org [2] Seurat. Integrated analysis of multimodal single-cell data. Cell. Yuhan Hao et. al.