vignette_seurat_part.qmd

---
title: "Exprest vignette / Seurat example"
format: html
---

To demonstrate *exprest* with real data, we use the Blood dataset from the Tabula Sapiens (Tabula Sapiens Consortium, [Science, 2022](https://doi.org/10.1126/science.abl4896)). This data set is available as AnnData file (h5ad format) from the [FigShare data page](https://figshare.com/articles/dataset/Tabula_Sapiens_v2/27921984).

```{r message=FALSE}
library(tidyverse)
library(Seurat)
```

We load the AnnData object in the h5ad file using the [schard](https://github.com/cellgeni/schard) package, which converts it to a Seurat object while loading.

```{r}
seu <- schard::h5ad2seurat("~/Blood_TSP1_30_version2d_10X_smartseq_scvi_Nov122024.h5ad")
```

The data set contains data for 85k cells.

Here is the UMAP plot with the cell-type annotations as provided in the object:

```{r}
DimPlot(seu, reduction="Xumap_", group.by = "free_annotation", label=FALSE) + 
  coord_equal() + guides(color = "none")
```

Here with colour guide:

```{r}
DimPlot(seu, reduction="Xumap_", group.by = "free_annotation") + coord_equal() 
```

We pick one gene, LCP1.

```{r}
gene_id <- "LCP1"

seu@assays$RNA@meta.features[gene_id,]
```

LCP is a good example as it has expression in several of the cell types but at different level:

```{r}
FeaturePlot(seu, gene_id, reduction="Xumap_") + coord_equal()
```

```{r warning=FALSE,fig.width=7,fig.height=6.5}
VlnPlot(seu, gene_id, alpha=.07, group.by="free_annotation") + guides(fill = "none")
```