I am currently starting to work, for the first time, with 450k methylation array data. I started by loading the IlluminaHumanMethylation450k.db package, but then noted that it has been deprecated in favour of the new package FDb.InfiniumMethylation.hg19, which seems to contain, instead of the usual AnnotationDb objects, an object of class FeatureDB, defined in the GenomicFeatures package.
Do we have some documentation on this new class? The help pages for FeatureDb and features are rather thin, and the package vignette only explains the TranscriptDB class but does not mention FeatureDB.
What I want to do is simple: I have a vector of Illumina CpG identifiers (like "cg20253340") and would like to get a GRanges object with their location.
I managed to do this with the old IlluminaHumanMethylation450k.db package but as it is deprecated and I am only about to start my project, I should probably use the new FDb.InfiniumMethylation.hg19 package. Should I, or is this all still too new for production use?
Is this the general idea of featureDB objects that they have one simple function that produces a table? I guess there is more to them. Although I don't want to complain: the table is, in a way, all one might ever need.
The idea for these particular DB packages is to encapsulate the two different Illumina Methylation platforms in a single package, and allow you to extract data for whatever platform(s) you care about.
So you can either extract everything using getPlatform(), or subsets thereof using the other two functions.
The FeatureDb class is just a generic container for storing genomic locations of arbitrary genomic features, in a SQLite data base. For example a TranscriptDb is a subclass of FeatureDb for storing transcript information.
So there is no idea that there would be a simple function per se. It all depends on what you put in it, and how you envision people using it. In this situation, three functions pretty much covers all the relevant use cases.
I should note that get450k() comes after library(FDb.InfiniumMethylation.hg19).
Best,
Jim
Okay, that is simple.
Is this the general idea of featureDB objects that they have one simple function that produces a table? I guess there is more to them. Although I don't want to complain: the table is, in a way, all one might ever need.
Thanks.
Simon
The idea for these particular DB packages is to encapsulate the two different Illumina Methylation platforms in a single package, and allow you to extract data for whatever platform(s) you care about.
So you can either extract everything using getPlatform(), or subsets thereof using the other two functions.
The FeatureDb class is just a generic container for storing genomic locations of arbitrary genomic features, in a SQLite data base. For example a TranscriptDb is a subclass of FeatureDb for storing transcript information.
So there is no idea that there would be a simple function per se. It all depends on what you put in it, and how you envision people using it. In this situation, three functions pretty much covers all the relevant use cases.
Best,
Jim