In AlfonsoRReyes/StudentGWAS:

Pre-requisites

Install these packages:

install.packages("RSQLite")
install.packages("ncdf4")

2 Part I: Implementing the GWASdata class

The process of designing and implementing a new class requires that the devel- opper spends some time thinking about:

• what s/he wants to achieve exactly with the class,
• how is the class going to be used, by who, for doing what,
• what are the typical use cases,
• what is the typical size of the data that will be manipulated, small (< 1 Mb), big (> 100 Mb), very big (> 10 Gb),
• how the class will interact with other packages and classes in CRAN/Bioconductor,
• how the facilities provided by the class will fit within the tools and file formats commonly used inside or outside Bioconductor,
• etc...

It’s generally considered good design to avoid storing redundant information (although some exceptions can be made for performance considerations) and to keep things as simple as possible.

2.1 Class definition

For our GWASdata class, we want the following slots:

• datapath: the path to a NetCDF file containing the GWAS experimental data (a matrix). This will be of type character.

• dataconn: the connection to the NetCDF file pointed by datapath. A connection to a NetCDF file is an object of class ncdf . Note that there is a complication here due to the fact that the ncdf class is defined in the ncdf package and that this package has no NAMESPACE. This makes it impossible to import the ncdf package in our StudentGWAS package, and, in particular, we can’t use "ncdf" to specify the type of our slot. So we will use the type "list" for this slot. This works because ncdf is an S3 class (aka old style class) and in the S3 class system, every object is a list with a class attribute attached to it.

• metadatapath: the path to an SQLite db file containing the GWAS meta- data. This will be of type character.

• metadataconn: the connection to the SQLite db pointed by metadatap- ath. This will be of type SQLiteConnection.

• nrow: the nb of rows in the GWAS matrix (this is also the nb or rows in the subjects table contained in the SQLite db). This will be of type integer.

• ncol: the nb of cols in the GWAS matrix (this is also the nb or rows in the snps table contained in the SQLite db). Also of class integer.

Exercise 1

Start a new file (let’s name it GWASdata-class.R) and write the setClass state ment for the GWASdata class.

  setClass("GWASdata",
  representation(
  datapath="character",
  ...
  ...
  )
  )

datapath <- file.path("./inst/extdata", "small_snpData.nc")
datapath

metadatapath <- file.path("./inst/extdata", "small_metadata.sqlite")
metadatapath

pth <- file.path("../inst/extdata", "snpdata.csv")
dat <- scan(pth, what = character(0), sep =",",
                               skip = 600, nlines = 1, quiet = TRUE)

library(RSQLite)
library(RNetCDF)


datapath <- file.path("../inst/extdata", "small_snpData.nc")
metadatapath <- file.path("../inst/extdata", "small_metadata.sqlite")

source("../R/GWASdata-class.R")
GWASdataRead(datapath, metadatapath)

source("../R/GWASdata-class.R")
datapath <- file.path("../inst/extdata", "small_snpData.nc")
metadatapath <- file.path("../inst/extdata", "small_metadata.sqlite")

gwas <- GWASdata(datapath = datapath, metadatapath = metadatapath)

dataconn(gwas)

nrow(gwas)

ncol(gwas)

AlfonsoRReyes/StudentGWAS documentation built on May 5, 2019, 4:54 a.m.