---
title: "CENTREprecomputed"
author: "Sara Lopez Ruiz de Vargas"
date: "`r doc_date()`"
vignette: >
  %\VignetteIndexEntry{CENTRE ExperimentHub datasets}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteDepends{}
output: 
  BiocStyle::html_document
link-citations: true
---

# CENTREprecomputed: An Experiment Data package interface with the precomputed 
data and example data for CENTRE

**CENTRE** is a package for Cell-type specific ENhancer Target pREdiction, that 
follows this workflow: 

`createPairs()` -> `computeGenericFeatures()` -> `computeCellTypeFeatures()` -> 
`centreClassification()`

The step `CENTRE::computeGenericFeatures()` computes the genomic distance and 
gets the precomputed features saved in the PrecomputedDataLight.db 
SQLite database. Visit `vignette("Centre-vignette")` for more information.

All of the data in the CENTREprecomputed package can be accessed through 
ExperimentHub:

```{r load hub, results='hide'}
library(ExperimentHub, quietly = TRUE)
```

```{r setup}
hub <- ExperimentHub()
eh <- query(hub, "CENTREprecomputed")
eh
```

The records EH9541-EH9545 are the example and test data for the CENTRE package.
Record EH9540 is the SQLite database which is accessed through `eh[["EH9540"]]` 
and returns a CENTREprecompDb object.

## The CENTREprecompDb object

The R object to represent the precomputed database is `CENTREprecompDb`. 
To get the slots for class run: 
```{r CENTREprecompDb object}
library(CENTREprecomputed)
centreprecompdb <- eh[["EH9540"]]
```

There are three tables inside the database: 

- `combinedTestData`: -log transformed p-values of the Wilcoxon rank sum tests.
- `crup_cor`: correlations of CRUP-EP scores and CRUP-PP scores across 
cell-types.
- `metadata`: metadata for the ExperimentHub

For more information check `vignette("Centre-vignette")` or the 
[CENTRE publication](https://academic.oup.com/bioinformatics/article/39/11/btad687/7429396?searchresult=1) 

The database can be used as follows: 
```{r how to fetch data}
# get all tables and their columns, won't show metadata table
tables(centreprecompdb)
# Select all cor_CRUP coefficients of enhancer EH38E3440167. Return the pair
# and ID column
fetch_data_precomp(centreprecompdb,
    table = "crup_cor",
    columns = c("pair", "cor_CRUP"),
    entries = "EH38E3440167",
    column_filter = "symbol38"
)
```

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