In Kayla-Morrell/HubGrub: A package to search and retrieve data from Hub resources

knitr::opts_chunk$set(
    collapse = TRUE,
    comment = "#>"
)

Introduction

HubGrub is a package that provides users with functionality to discover data in the Bioconductor Hub structures. The package helps the user discover what type of data is available for retrieval and can display the metadata of that data. There is also functionality to retrieve the data and import it without needing to download the entire file. We will demonstrate all the functionality of the package in the following vignette.

Installation

Install the most recent version from Bioconductor:

if(!requireNamesapce("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("HubGrub")

The development version is also available for install from GitHub:

BiocManager::install("Kayla-Morrell/HubGrub")

Then load HubGrub:

library(HubGrub)

HubGrub

Discover data in Hubs

The first step when looking into the hubs is to figure out what type of data is available. Once a hub object is created, either from AnnotationHub or ExperimentHub, then the user can use discoverData() to see how many resources are available for a specific data type. For example, say we want to find out how many bigwig files are in Bioconductor's AnnotationHub.

library(AnnotationHub)
ah = AnnotationHub()
discoverData(ah, "BigWigFile")

Info on data in Hubs

Once we figure out which files we are interested in we can use dataInfo() to see the metadata information on these files. This is where the user can narrow down their search to specific files they are interested in retrieving.

tbl <- dataInfo(ah, "BigWigFile")
tbl

A user maybe intersted in specifically human files of a specific cell type. Then they can use dplyr functionality on the table to create the subset of files they are interested in.

tbl <- tbl |> 
    dplyr::filter(species == "Homo sapiens") |> 
    dplyr::select(ID, title)
tail(tbl)

Importing the files

With a subset of files that the user is interested in, importData() will read in a part of the file without having to download the entire resource. These files can be quite large, so importing just a portion of a file can be beneficial to users.

In order to use this function to it's fullest potential, the user should define a range of data that they are interested in. This range of data should be in the form of a GRanges object.

library(GenomicRanges)
which <- GRanges(c("chr2", "chr2"), IRanges(c(1, 300), c(400, 1000000)))
importData(ah, "AH49544", which)

Session Information

sessionInfo()

Kayla-Morrell/HubGrub documentation built on Dec. 18, 2021, 2:43 a.m.