---
title: "scRNASeq HNSC data using Bioconductor's ExperimentHub"
author: "Mariano J. Alvarez"
date: "`r doc_date()`"
vignette: >
  %\VignetteIndexEntry{scRNASeq HNSC data using Bioconductor's ExperimentHub}
  %\VignetteEngine{knitr::rmarkdown}
output: 
  BiocStyle::html_document
---


# Accessing human HNSC scRNASeq data using Bioconductor's ExperimentHub

Transcripts per million (TPM) single cell RNA-Seq data for 5,902 cells from
18 patients--oral cavity head and neck squamous cell carcinoma (HNSC)--
are available from GEO
[GSE103322](http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE103322).
These data are also available as a *SingleCellExpression* from ExperimentHub.

In the example below, we show how this dataset can be dwnloaded from 
ExperimentHub.

```{r get-sce}
library(ExperimentHub)
library(SingleCellExperiment)
eh = ExperimentHub()
dset <- query(eh , "GSE103322")
dset
```
One can then extract the data for this using
```{r download-sce}
sce <- dset[[1]]
```

##  Exploring the metadata

The metadata is available from the *SingleCellExpression* object with

```{r metadata}
head(SummarizedExperiment::colData(sce))
```

For example, to obtain the number of cells classified as non-tumor types

```{r non-tumor}
table(SummarizedExperiment::colData(sce)$non.cancer.cell.type)
```

##  Extracting the data 

The data can be extracted from the *SingleCellExpression* object with

```{r data}
dset <- SummarizedExperiment::assays(sce)$TPM
dim(dset)
dset[1:4, 1:3]
```

# sessionInfo()

```{r sessionInfo}
sessionInfo()
```