---
title: "SCoPE2: macrophage vs monocytes"
date: "`r BiocStyle::doc_date()`"
vignette: |
    %\VignetteIndexEntry{SCoPE2: macrophage vs monocytes}
    %\VignetteEngine{knitr::rmarkdown}
    %\VignetteEncoding{UTF-8}
output:
    BiocStyle::html_document:
    toc_float: true
Package: SingleCellMultiModal
bibliography: ../inst/REFERENCES.bib
---

This vignette will guide you through how accessing and manipulating 
the SCoPE2 data sets available from the `SingleCellMultimodal` package.
 
# Installation
    
```{r,eval=FALSE}
if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("SingleCellMultiModal")
```

## Load

```{r,include=TRUE,results="hide",message=FALSE,warning=FALSE}
library(SingleCellMultiModal)
library(MultiAssayExperiment)
```

# SCoPE2

SCoPE2 is a mass spectrometry (MS)-based single-cell proteomics 
protocol to quantify the proteome of single-cells in an untargeted
fashion. It was initially developed by @Specht2020-jm.

## Downloading data sets

The user can see the available data set by using the default options.

```{r}
SCoPE2("macrophage_differentiation", 
       mode = "*", 
       version = "1.0.0", 
       dry.run = TRUE)
```

Or by simply running:
    
```{r}
SCoPE2("macrophage_differentiation")
```

## Available projects

Currently, only the `macrophage_differentiation` is available.

## Retrieving data

You can retrieve the actual data from `ExperimentHub` by setting 
`dry.run = FALSE`. This example retrieves the complete data set 
(transcriptome and proteome) for the `macrophage_differentiation` 
project:
    
```{r,message=FALSE}
scope2 <- SCoPE2("macrophage_differentiation",
                 modes = "rna|protein",
                 dry.run = FALSE)
scope2
```

# The macrophage differentiation project

This data set has been acquired by the Slavov Lab (@Specht2020-jm). 
It contains single-cell proteomics and single-cell 
RNA sequencing data for macrophages and monocytes. The objective of the 
research that led to generate the data is to understand whether 
homogeneous monocytes differentiate in the absence of cytokines to
macrophages with homogeneous or heterogeneous profiles. The transcriptomic and 
proteomic acquisitions are conducted on two separate subset of similar 
cells (same experimental design). The cell type of the samples are known only 
for the **proteomics** data. The proteomics data was retrieved from 
the authors' [website](https://scope2.slavovlab.net/docs/data) and the
transcriptomic data was retrieved from the GEO database (accession id:
[GSE142392](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE142392)).

For more information on the protocol, see @Specht2020-jm.

## Data versions

Only version `1.0.0` is currently available.

The `macrophage_differentiation` data set in this package contains two
assays: `rna` and `protein`. 

### Cell annotation

The single-cell proteomics data contains cell type annotation 
(`celltype`), sample preparation batch (`batch_digest` and 
`batch_sort`), chromatography batch (`batch_chromatography`), and the 
MS acquisition run (`batch_MS`). The single-cell transcriptomics data 
was acquired in two batches (`batch_Chromium`). Note that because the 
cells that compose the two assays are distinct, there is no common 
cell annotation available for both proteomics and transcriptomics. The
annotation were therefore filled with `NA`s accordingly.

```{r}
colData(scope2)
```

### Transcriptomic data 

You can extract and check the transcriptomic data through subsetting:

```{r}
scope2[["rna"]]
```

The data is rather large and is therefore stored on-disk using the 
HDF5 backend. You can verify this by looking at the assay data matrix.
Note that the counts are UMI counts.

```{r}
assay(scope2[["rna"]])[1:5, 1:5]
```

### Proteomic data 

The `protein` assay contains MS-based proteomic data. 
The data have been passed sample and feature quality control, 
normalized, log transformed, imputed and batch corrected. Detailed 
information about the data processing is available in 
[another vignette](https://uclouvain-cbio.github.io/SCP.replication/articles/SCoPE2.html). You can extract the proteomic data similarly to the 
transcriptomic data:

```{r}
scope2[["protein"]]
```

In this case, the protein data have reasonable size and are loaded
directly into memory. The data matrix is stored in `logexprs`. We 
decided to not use the traditional `logcounts` because MS proteomics 
measures intensities rather than counts as opposed to scRNA-Seq. 

```{r}
assay(scope2[["protein"]])[1:5, 1:5]
```

# sessionInfo

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

# References