<!--
%\VignetteEngine{knitr::knitr}
%\VignetteIndexEntry{Salmon quantifications for human macrophage immune response experiment}
-->

# Experiment and sample description

This package provides the output of running Salmon on a set of 24
RNA-seq samples from Alasoo, et al. "Shared genetic effects on
chromatin and gene expression indicate a role for enhancer priming in
immune response", published in Nature Genetics, January 2018 
[doi: 10.1038/s41588-018-0046-7](https://doi.org/10.1038/s41588-018-0046-7).

6 donors were selected from those that had publicly available reads in
ENA. The selection procress is recorded in the file
`inst/scripts/ena_download.R`, the donors were chosen as all female,
the sample having been received as frozen, and then the top 6 were
chosen based on mean purity as recorded by the study authors.

# Salmon quantification

[Salmon](https://combine-lab.github.io/salmon/)
version 0.12.0 was run using 
[Gencode](https://www.gencodegenes.org/)
human reference transcripts v29, with a 
[snakemake](https://snakemake.readthedocs.io/en/stable/) 
script that can be seen in `inst/scripts/Snakemake` and with 
log output in `inst/scripts/snakemake.log`. 

The quantification output is in the `inst/extdata/quants`
directory. The `quant.sf` files have been gzipped to preserve space,
so they are stored as `quant.sf.gz`. Salmon was run with 20 Gibbs
inferential replicates for each sample. To reduce the size of the
package, the inferential replicates (`aux_info/bootstrap`) and the
`aux_info/ambig_info.tsv` files are no longer included in
`inst/extdata/quants` (as of version 1.29.1). The full Salmon output,
including the inferential replicates, is archived on Zenodo:
[doi: 10.5281/zenodo.22982581](https://doi.org/10.5281/zenodo.22982581).
The inferential replicates for a subset of genes are provided in the
`macro_txp_se` object (see below).

The file `inst/extdata/gencode.v29.annotation.slim.gtf.gz` is a
reduced version of the Gencode v29 GTF file, containing only the
`gene` and `transcript` features, with a subset of the attributes
(`gene_id`, `transcript_id`, `gene_type`, `gene_name`,
`transcript_type`, `transcript_name`). It can be used to build
transcript and gene ranges, but not exon ranges. It was created with
the script `inst/scripts/makeSlimGTF.sh`.

# Sample information

The `inst/scripts/ena_downloads.R` R script also shows how the column
data file `coldata.csv` was generated. This file contains the
phenotypic information about the 24 samples:

```{r}
dir <- system.file("extdata", package="macrophage")
coldata <- read.csv(file.path(dir,"coldata.csv"))
coldata <- coldata[,c(1,2,3,5)]
coldata
```

# SummarizedExperiment

The package also contains a summarized experiment object created
using the script in `inst/scripts/gse_create.R`. This object can be
loaded with `data("gse")`.

Additionally, the package contains a transcript-level
*SummarizedExperiment*, `macro_txp_se`, including the 20 inferential
replicates, for all transcripts of a random subset of 2000 expressed
genes plus the GBP gene family, and a transcript-to-gene table,
`macro_tx2gene`. These were created with the script in
`inst/scripts/makeSumExp.R` and are used in the *fishpond* vignette.
See `?macro_txp_se` and `?macro_tx2gene` for details.

# Acknowledgments 

Thanks to the study authors for posting their data publicly and
clearly labelling their data.

# Session info

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