libpackages.Rmd

# Library and packages

* **Packages** are collections of R functions, data, and compiled code in a well-defined format.

* The directory where packages are stored is called the **library**.

*Source of definitions: http://www.statmethods.net/interface/packages.html*

## R base

A set a standard packages which are supplied with R by default.<br>
Example: package **base** (write, table, rownames functions), package **utils** (read.table, str functions), package **stats** (var, na.omit, median functions).

## R contrib

All other packages:

* [CRAN](https://cran.r-project.org): Comprehensive R Archive Network
  + 17318<sup>*</sup> packages available
  + find packages in https://cran.r-project.org/web/packages/
	<img src="images/cran_packages.png" width="550"/>
* [Bioconductor](https://www.bioconductor.org/):
	+ 3372<sup>*</sup> packages available
	+ find packages in https://bioconductor.org/packages
	<img src="images/bioc_packages.png" width="550"/>

*<sup>*</sup>As of March 2021*

<h4>Bioconductor</h4>

Set of R packages specialized in the analysis of bioinformatics data.<br>

Bioconductor supports most types of **genomics and NGS data** (e.g. limma, DESeq2, BayesPeak) and integrates:

* Specific data classes (e.g. Granges from GenomicRanges)
* Integrates command line tools (e.g Rsamtools)
* Annotation tools (e.g. biomaRt)

There are different types of Bioconductor packages:

* **Software**: set of functions
  + e.g. DESeq2 (NGS data analysis)
* **Annotation**: annotation of specific arrays, organisms, events, etc.
  + e.g. BSgenome.Hsapiens.UCSC.hg38
* **Experiment**: data that can be loaded and used
  + e.g. ALL (acute lymphoblastic leukemia dataset)

## Install a package

* With RStudio, from the bottom-right panel, "Packages" tab:

<img src="images/rstudio_installBiocManager_arrow.png" width="550"/>

* From the console:
```{r, eval=F}
install.packages(pkgs="BiocManager")
```

* Install a bioconductor package:
```{r, eval=F}
# Install Bioconductor package manager
install.packages(pkgs="BiocManager")
# Install Bioconductor package
BiocManager::install("DESeq2")
```


## Load a package

* With RStudio: tick the box.

<img src="images/rstudio_tick_BiocManager.png" width="450"/>

* From the console:
```{r}
library("BiocManager")
```

## Check what packages are currently loaded

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

## List functions from a package

* With RStudio: click on the package name in the "Packages" tab: <br>

<img src="images/rstudio_list_biocmanager.png" width="450"/>

* From the console
```{r, eval=F}
ls("package:BiocManager")
```

<br>

**HANDS-ON**

1. Install and load package **WriteXLS** (using either the RStudio "Packages" panel, or the console)

2. Create a simple matrix (of your choice) with 6 rows and 4 columns!

3. Write matrix to a file using the **WriteXLS** function (from the package of the same name). Check the help page of the function!


## RStudio server at CRG

If you can't install packages (permission issues), you first need to specify a writeable directory to install the packages into.<br>

Follow the steps below:

```{r, eval=F}
# Go to your home directory
setwd("~")
# Create a directory where to store the packages
dir.create("R_packages")
# Add directory location to the library path
.libPaths("~/R_packages/")
```