%\VignetteIndexEntry{Using the inSilicoDb v2 package} 
%\VignetteDepends{RCurl, rjson} 

\documentclass{article}

\usepackage{url}

\usepackage{color}
\newcommand{\todo}[1]{\textcolor{red}{\textbf{#1}}}

\begin{document}

\title{Using inSilicoDb 2.0}
\author{Quentin De Clerck$\footnote{\texttt{qdeclerc@vub.ac.be}}$}
\maketitle

\section{Functions from inSilicoDb 2.0}
This new version of the package provides all the functionality of
previous inSilicoDb packages developed by \emph{Jonatan Taminau}. The functions
did not change, therefore we refer to the description of previous versions.

\subsection{Access to InSilico MySafe}
One of the new features of this package is the possibility for users to access
their private data stored on InSilico MySafe. This feature is implemented by the
functions:
\begin{description}
  \item[InSilicoLogin(login, password)] Logs the user in for the given login
  and password. There is no secure way to login to a webservice in R. Therefore
  the password given to the function has to be the md5 hash of the real
  password. After login in, the user can use the normal functions as described
  earlier with their private data.
  \item[InSilicoLogout()] Logs the currently user out of the InSilico DB
  webservice.
  \item[getInSilicoUserDetails()] Returns informations (id, name, email) about
  the user that is currently logged in.
\end{description}


\subsection{Check the accessibility of data}
Some helper functions were added to provide information about the availability
of data: getDatasetInfo and getPlatformList.

The purpose of getDatasetInfo is two-fold. First, it will return the default
values of all optional parameters and the title of the study. Second, it will
return the availability of the requested dataset for specified parameters. It
returns an error if the data is not available for download.
You need to login to access datasets and datasetinfo. Use your InSilicoDB login and an md5 hash of your password. For this example we're using a restricted test account.
<<getDatasetInfo>>=

library("inSilicoDb");
InSilicoLogin("rpackage_tester@insilicodb.com", "5c4d0b231e5cba4a0bc54783b385cc9a");

eset = getDatasetInfo("GSE781", "GPL96");
print(eset);

## We check the availability of following normalizations for series GSE781 on platform GPL97
norms = c("FRMA", "ORIGINAL")
output = sapply(norms, function(n) {
	tryCatch({
			eset <- getDatasetInfo("GSE781", "GPL97", norm = n);
			eset$norm;
		}, error = function(e) {
			"Unavailable"
		});
	});

print(output);

# We can thus conclude that the series GSE781 on platform GPL97 does not support FRMA
@

The getPlatformList function returns all the platforms supported by the package.
It is also possible to query the availability of the normalizations using this
function.

<<getPlatformList>>=
platforms = getPlatformList();

print(platforms);

FRMAplatforms = getPlatformList(norm = "FRMA");

print(FRMAplatforms);

@

\subsection{Curated Data}
What changed in the functionality with respect to previous
package is the better integration of the curated data. The user can now decide if he
wants curated data or not by specifying the format.There are two
possible formats: ESET and CURESET. By default the format is CURESET, this
means with curated data. It is also possible to specify the curation wanted for
the requested expression set. 

<<curated_dataset>>=
# without curated data
eset = getDataset("GSE4635", "GPL96", format = "ESET"); 
# with curated data
cureset = getDataset("GSE4635", "GPL96", format = "CURESET", curation = 9016);

print(phenoData(eset));
print(phenoData(cureset));
@
It is possible to get information about the default curation (id, time, curator)
of a particular dataset by calling the getDefaultCuration function. Using the
getCurationInfo function returns a complete overview of all the possible
curations in more details.

<<getCurationInfo>>=
default = getDefaultCuration("GSE4635");

print(default);


getCurationInfo("GSE4635");

@

\section{SCAN and UPC normalizations}

Two normalizations were added in to the package. It is now possible to request
data in Single Channel Array Normalization (SCAN) \cite{piccolo2013scan} or
Universal exPression Codes (UPC) \cite{piccolo2013upc}. SCAN is a possible
alternative to fRMA and UPC outputs the probabilities of the probes/genes being expressed.

<<SCAN_and_UPC>>=
scan = getDataset("GSE7670", "GPL96", norm = "SCAN", features = "gene");
# example of values in SCAN normalization
print(exprs(scan)[1:10, 1:5]);
upc = getDataset("GSE7670", "GPL96", norm = "UPC", features = "gene");
# example of values in UPC normalization
print(exprs(upc)[1:10, 1:5]);
@

\section{Session Info}
<<>>=
sessionInfo()
@


\bibliographystyle{plain}
\bibliography{inSilicoDb2}
\end{document}