r-bioc-annotationdbi 1.26.1-1 / usr / lib / R / site-library / AnnotationDbi / doc / IntroToAnnotationPackages.Rnw

%\VignetteIndexEntry{AnnotationDbi: Introduction To Bioconductor Annotation Packages}
%\VignetteDepends{org.Hs.eg.db,TxDb.Hsapiens.UCSC.hg19.knownGene,hom.Hs.inp.db}
%\VignetteKeywords{annotation, database}
%\VignettePackage{AnnotationDbi}
%\VignetteEngine{knitr::knitr}

\documentclass[11pt]{article}

<<style, eval=TRUE, echo=FALSE, results='asis'>>=
BiocStyle::latex()
@

%% Question, Exercise, Solution
\usepackage{theorem}
\theoremstyle{break}
\newtheorem{Ext}{Exercise}
\newtheorem{Question}{Question}


\newenvironment{Exercise}{
  \renewcommand{\labelenumi}{\alph{enumi}.}\begin{Ext}%
}{\end{Ext}}
\newenvironment{Solution}{%
  \noindent\textbf{Solution:}\renewcommand{\labelenumi}{\alph{enumi}.}%
}{\bigskip}


\title{AnnotationDbi: Introduction To Bioconductor Annotation Packages}
\author{Marc Carlson}

%% \SweaveOpts{keep.source=TRUE}

<<include=FALSE>>=
library(knitr)
opts_chunk$set(tidy=FALSE)
@


\begin{document}

\maketitle


\begin{figure}[ht]
\centering
\includegraphics[width=.6\textwidth]{databaseTypes.pdf}
\caption{Annotation Packages: the big picture}
\label{fig:dbtypes}
\end{figure}

\Bioconductor{} provides extensive annotation resources. These can be
\emph{gene centric}, or \emph{genome centric}. Annotations can be
provided in packages curated by \Bioconductor, or obtained from
web-based resources.  This vignette is primarily concerned with
describing the annotation resources that are available as
packages. More advanced users who wish to learn about how to make new
annotation packages should see the vignette titled "Creating select
Interfaces for custom Annotation resources" from the
\Rpackage{AnnotationForge} package.

Gene centric \Rpackage{AnnotationDbi} packages
include:
\begin{itemize}
  \item Organism level: e.g. \Rpackage{org.Mm.eg.db}.
  \item Platform level: e.g. \Rpackage{hgu133plus2.db},
        \Rpackage{hgu133plus2.probes}, \Rpackage{hgu133plus2.cdf}.
  \item Homology level: e.g. \Rpackage{hom.Dm.inp.db}.
  \item System-biology level: \Rpackage{GO.db}
\end{itemize}
Genome centric \Rpackage{GenomicFeatures} packages include
\begin{itemize}
  \item Transcriptome level: e.g. \Rpackage{TxDb.Hsapiens.UCSC.hg19.knownGene}
  \item Generic genome features: Can generate via \Rpackage{GenomicFeatures}
\end{itemize}
One web-based resource accesses
\href{http://www.biomart.org/}{biomart}, via the \Rpackage{biomaRt}
package:
\begin{itemize}
  \item Query web-based `biomart' resource for genes, sequence,
        SNPs, and etc.
\end{itemize}


The most popular annotation packages have been modified so that they
can make use of a new set of methods to more easily access their
contents.  These four methods are named: \Rfunction{columns},
\Rfunction{keytypes}, \Rfunction{keys} and \Rfunction{select}.  And they are
described in this vignette.  They can currently be used with all chip,
organism, and \Rclass{TranscriptDb} packages along with the popular
GO.db package.

For the older less popular packages, there are still conventient ways
to retrieve the data.  The \emph{How to use bimaps from the ".db"
annotation packages} vignette in the \Rpackage{AnnotationDbi} package
is a key reference for learnign about how to use bimap objects.

Finally, all of the `.db' (and most other \Bioconductor{} annotation
packages) are updated every 6 months corresponding to each release of
\Bioconductor{}.  Exceptions are made for packages where the actual
resources that the packages are based on have not themselves been
updated.


\subsection{AnnotationDb objects and the select method}

As previously mentioned, a new set of methods have been added that
allow a simpler way of extracting identifier based annotations.  All
the annotation packages that support these new methods expose an
object named exactly the same as the package itself.  These objects
are collectively called \Rclass{AnntoationDb} objects for the class
that they all inherit from.  The more specific classes (the ones that
you will actually see in the wild) have names like \Rclass{OrgDb},
\Rclass{ChipDb} or \Rclass{TranscriptDb} objects.  These names
correspond to the kind of package (and underlying schema) being
represented.  The methods that can be applied to all of these objects
are \Rfunction{columns}, \Rfunction{keys}, \Rfunction{keytypes} and
\Rfunction{select}.


\subsection{ChipDb objects and the select method}

An extremely common kind of Annotation package is the so called
platform based or chip based package type.  This package is intended
to make the manufacturer labels for a series of probes or probesets to
a wide range of gene-based features.  A package of this kind will load
an \Rclass{ChipDb} object.  Below is a set of examples to show how you
might use the standard 4 methods to interact with an object of this
type.

First we need to load the package:

<<loadChip>>=
library(hgu95av2.db)
@ 

If we list the contents of this package, we can see that one of the
many things loaded is an object named after the package "hgu95av2.db":

<<listContents>>=
ls("package:hgu95av2.db")
@ 

We can look at this object to learn more about it:

<<show>>=
hgu95av2.db
@ 

If we want to know what kinds of data are retriveable via
\Rfunction{select}, then we should use the \Rfunction{columns} method like
this:

<<columns>>=
columns(hgu95av2.db)
@ 

If we are further curious to know more about those values for columns, we
can consult the help pages.  Asking about any of these values will
pull up a manual page describing the different fields and what they
mean.

<<help, eval=FALSE>>=
help("SYMBOL")
@ 

If we are curious about what kinds of fields we could potentiall use
as keys to query the database, we can use the \Rfunction{keytypes}
method.  In a perfect world, this method will return values very
similar to what was returned by \Rfunction{columns}, but in reality, some
kinds of values make poor keys and so this list is often shorter.

<<keytypes>>=
keytypes(hgu95av2.db)
@ 

If we want to extract some sample keys of a particular type, we can
use the \Rfunction{keys} method.

<<keys>>=
head(keys(hgu95av2.db, keytype="SYMBOL"))
@ 

And finally, if we have some keys, we can use \Rfunction{select} to
extract them.  By simply using appropriate argument values with select
we can specify what keys we want to look up values for (keys), what we
want returned back (columns) and the type of keys that we are passing in
(keytype)

<<selectChip>>=
#1st get some example keys
k <- head(keys(hgu95av2.db,keytype="PROBEID"))
# then call select
select(hgu95av2.db, keys=k, columns=c("SYMBOL","GENENAME"), keytype="PROBEID")
@ 

And as you can see, when you call the code above, select will try to
return a data.frame with all the things you asked for matched up to
each other.

\subsection{OrgDb objects and the select method}
An organism level package (an `org' package) uses a central gene
identifier (e.g. Entrez Gene id) and contains mappings between this
identifier and other kinds of identifiers (e.g. GenBank or Uniprot
accession number, RefSeq id, etc.).  The name of an org package is
always of the form \emph{org.<Ab>.<id>.db}
(e.g. \Rpackage{org.Sc.sgd.db}) where \emph{<Ab>} is a 2-letter
abbreviation of the organism (e.g. \Rpackage{Sc} for
\emph{Saccharomyces cerevisiae}) and \emph{<id>} is an abbreviation
(in lower-case) describing the type of central identifier
(e.g. \Rpackage{sgd} for gene identifiers assigned by the
Saccharomyces Genome Database, or \Rpackage{eg} for Entrez Gene ids).

Just as the chip packages load a \Rclass{ChipDb} object, the org
packages will load a \Rclass{OrgDb} object.  The following exercise
should acquaint you with the use of these methods in the context of an
organism package.

%% select exercise
\begin{Exercise}
  Display the \Rclass{OrgDb} object for the \Biocpkg{org.Hs.eg.db} package.
  
  Use the \Rfunction{columns} method to discover which sorts of
  annotations can be extracted from it. Is this the same as the result
  from the \Rfunction{keytypes} method?  Use the \Rfunction{keytypes}
  method to find out.  
  
  Finally, use the \Rfunction{keys} method to extract UNIPROT
  identifiers and then pass those keys in to the \Rfunction{select}
  method in such a way that you extract the gene symbol and KEGG pathway
  information for each.  Use the help system as needed to learn which
  values to pass in to columns in order to achieve this.
\end{Exercise}
\begin{Solution}
<<selectOrg1>>=
library(org.Hs.eg.db)
columns(org.Hs.eg.db)
@ 
<<selectOrg2, eval=FALSE>>=
help("SYMBOL") ## for explanation of these columns and keytypes values
@
<<selectOrg3>>=
keytypes(org.Hs.eg.db)
uniKeys <- head(keys(org.Hs.eg.db, keytype="UNIPROT"))
cols <- c("SYMBOL", "PATH")
select(org.Hs.eg.db, keys=uniKeys, columns=cols, keytype="UNIPROT")
@
\end{Solution}

So how could you use select to annotate your results? This next
exercise should hlep you to understand how that should generally work.

\begin{Exercise}
  Please run the following code snippet (which will load a fake data
  result that I have provided for the purposes of illustration):
  
<<selectData>>=
load(system.file("extdata", "resultTable.Rda", package="AnnotationDbi"))
head(resultTable)
@

  The rownames of this table happen to provide entrez gene identifiers
  for each row (for human).  Find the gene symbol and gene name for each
  of the rows in resultTable and then use the merge method to attach
  those annotations to it.
\end{Exercise}
\begin{Solution}
<<selectOrgData>>=
annots <- select(org.Hs.eg.db, keys=rownames(resultTable),
                 columns=c("SYMBOL","GENENAME"), keytype="ENTREZID")
resultTable <- merge(resultTable, annots, by.x=0, by.y="ENTREZID")
head(resultTable)
@
\end{Solution}


\subsection{Using select with GO.db}

When you load the GO.db package, a \Rclass{GODb} object is also
loaded. This allows you to use the \Rfunction{columns}, \Rfunction{keys},
\Rfunction{keytypes} and \Rfunction{select} methods on the contents of the
GO ontology.  So if for example, you had a few GO IDs and wanted to know
more about it, you could do it like this:

<<selectGO>>=
library(GO.db)
GOIDs <- c("GO:0042254","GO:0044183")
select(GO.db, keys=GOIDs, columns="DEFINITION", keytype="GOID")
@



% This here is the point where I am planning to fork my talk to the GenomicFeatures vignette

%\subsection{Using select with HomDb packages}



\subsection{Using select with TranscriptDb packages}

A \Rclass{TranscriptDb} package (a 'TxDb' package) connects a set of genomic
coordinates to various transcript oriented features.  The package can
also contain Identifiers to features such as genes and transcripts,
and the internal schema describes the relationships between these
different elements.  All TranscriptDb containing packages follow a
specific naming scheme that tells where the data came from as well as
which build of the genome it comes from.


%% select exercise TranscriptDb
\begin{Exercise}
  Display the \Rclass{TranscriptDb} object for the 
  \Biocpkg{TxDb.Hsapiens.UCSC.hg19.knownGene} package.
  
  As before, use the \Rfunction{columns} and \Rfunction{keytypes} methods to
  discover which sorts of annotations can be extracted from it.
  
  Use the \Rfunction{keys} method to extract just a few gene identifiers and
  then pass those keys in to the \Rfunction{select} method in such a
  way that you extract the transcript ids and transcript starts for each.
\end{Exercise}
\begin{Solution}
<<selectTxDb>>=
library(TxDb.Hsapiens.UCSC.hg19.knownGene)
txdb <- TxDb.Hsapiens.UCSC.hg19.knownGene
txdb
columns(txdb)
keytypes(txdb)
keys <- head(keys(txdb, keytype="GENEID"))
cols <- c("TXID", "TXSTART")
select(txdb, keys=keys, columns=cols, keytype="GENEID")

@
\end{Solution}

As is widely known, in addition to providing access via the
\Rfunction{select} method, \Rclass{TranscriptDb} objects also provide
access via the more familiar \Rfunction{transcripts},
\Rfunction{exons}, \Rfunction{cds}, \Rfunction{transcriptsBy},
\Rfunction{exonsBy} and \Rfunction{cdsBy} methods.  For those who do
not yet know about these other methods, more can be learned by seeing
the vignette called: \emph{Making and Utilizing TranscriptDb Objects}
in the \Rpackage{GenomicFeatures} package.




The version number of R and packages loaded for generating the vignette were:

<<SessionInfo, echo=FALSE>>=
sessionInfo()
@




\end{document}