Hypo Cell Atlas (HCA)

The Hypo Cell Atlas (HCA) is a comprehensive platform for single-cell sequencing data of mammalian hypothalamus. It provides tools for data visualization, analysis, and exploration, enabling users to gain insights into the complex biology of this vital brain region.

Home

The Data Retrieval box enables users to search for datasets using article names, authors, journals, keywords, etc.

The homepage provides an overview of the platform's data, which includes four species: human, mouse, macaque, and marmoset. It also displays the total number of datasets and samples for each species. Click on a species card to jump to the visualization of the corresponding species.

Dataset

HCA datasets were collected from reputable sources, including NCBI GEO, China National GeneBank (CNGBdb), European Nucleotide Archive (ENA), and The Neuroscience Multi-omic Archive (NeMO).

Data Pre-processing and Analysis

• Raw data were processed using cellranger.
• The R package Seurat (v5.1.0) was used for data normalization, scaling, PCA (top 2000 variable genes), and clustering via the Louvain algorithm.
• For each dataset, we identified potential marker genes for each cluster by Seurat's Wilcoxon rank sum test and annotated these clusters with these marker genes.

Data Visualization

• Data visualization is powered by the R package plotly, allowing interactive 3D exploration of UMAP reduction results.
• Users can filter datasets by species or dataset features via the selectors on the left.
• After selecting a species and dataset, users can further refine their selection by choosing specific samples and cell types, enabling data visualization tailored to these specified criteria.

Gene

• Users can explore the expression of specific genes across cell types using dot plots and violin plots.
• Filters on the left include options for species, state, sex, and developmental stage, enabling detailed views of gene expression under chosen conditions.
• The platform also supports co-expression analysis of two genes.

Cross-Sex

The Cross-Sex analysis integrates mouse data with explicit sex information using the SCVI method and consists of two sections: Sex Differences and Cell-Cell Communication.

Sex Differences

1. Users can filter data by condition, stage, and cell type.
2. Results include:
   • Proportional comparison of cell types between sexes.
   • Volcano plots for differentially expressed genes (DEGs) in specific cell types.
   • GO and KEGG enrichment analysis of DEGs.

Cell-Cell Communication

1. Analyzed using the R package cellchat, with filtering options for condition and stage.
2. Clicking "Run" initiates analysis; results are emailed upon completion due to long processing times.
3. Results include:
   • Visualization of communication counts, intensity, and signaling patterns.
   • Interaction comparisons for two or three selected cell types.
   • Comparison of signaling pathways shared by male and female samples.
   • Male-female comparisons of receptor-ligand pairs between user-specified signaling (source) and target cell types.

Note: Cross-Sex analysis currently supports only mouse data.

Cross-Species

The Cross-Species analysis integrates data from human, rhesus macaque, and marmoset using the scANVI method. Homologous gene symbols were converted to human equivalents using biomaRt and formatted with scCustomize. It also includes two sections: Species Differences and Cell-Cell Communication.

Species Differences

1. Users can compare any two species (human, marmoset, or macaque) for a selected state and cell type.
2. Results include:
   • Proportional comparison of cell types between species.
   • Differentially expressed genes (DEGs).
   • GO and KEGG enrichment analysis of DEGs.

Cell-Cell Communication

1. Analyzed using cellchat with the human ligand-receptor database, as all gene symbols were converted to human equivalents.
2. Results are divided into four parts, similar to those in the Cross-Sex section:
   • Communication counts, intensity, and signaling patterns.
   • Interaction comparisons for selected cell types.
   • Pathway comparisons between species.
   • Cross-species comparisons of receptor-ligand pairs between user-specified signaling (source) and target cell types.

Note: Cross-species analysis currently only supports normal condition data.

Download

• Each sample's raw data was pre-processed upstream using cellranger, and all curated datasets are stored in a standardized Seurat format (.rds), available for download.
• The table provides key information about the datasets, including species, sex, developmental stage, sequencing technology, and file name.
• A search box at the top of the table allows users to search for datasets using keywords.