%\VignetteIndexEntry{Analysis of alternative splicing using ASpli}
%\VignetteKeywords{Alternative splicing analysis}
%\VignettePackage{ASpli}
\documentclass{article}
<<style, echo=FALSE, results=tex>>=
BiocStyle::latex()
options(continue=" ")
@

\usepackage{verbatim} 
\usepackage{caption} 
\usepackage{cite}
\usepackage{subcaption}
\usepackage{hyperref}
\usepackage{amsmath}
\usepackage{float}

\newcommand{\RNAseq}{\textrm{\textbf{RNA}-\textbf{Seq}}}
\newcommand{\secref}[1]{\ref{#1} : \nameref{#1}}

\newfloat{captionedEq}{thp}{eqc}
\floatname{captionedEq}{Equation}

\title{\texttt{ASpli}: An integrative R package for analysing alternative 
  splicing using \RNAseq}

\author{Estefania Mancini, Javier Iserte, Marcelo Yanovsky, Ariel Chernomoretz}

\begin{document}

\SweaveOpts{concordance=TRUE}

\maketitle

\tableofcontents

\section{Introduction}

Alternative splicing (AS) is a common mechanism of post-transcriptional gene 
regulation in eukaryotic organisms that expands the functional and regulatory 
diversity of a single gene by generating multiple mRNA isoforms that encode 
structurally and functionally distinct proteins. The development of novel 
high-throughput sequencing methods for RNA (\RNAseq) has provided a powerful 
mean of studying AS under multiple conditions and in a genome-wide manner 
\cite{pmid24549677}. However, using \RNAseq to study changes in AS under 
different experimental conditions is not trivial. 
In this vignette, we describe how to use ASpli, an integrative and user-friendly
R package that facilitates the analysis of changes in both annotated and novel 
AS events. This package combines statistical information from exon, intron, and 
splice junction differential usage (p-value, FDR), with information from splice 
junction reads to calculate differences in the percentage of exon inclusion 
($\Delta$PSI) and intron retention ($\Delta$PIR). The proposed methodology 
reliably reflect the magnitude of changes in the relative abundance of different 
annotated and novel AS events. This method can be used to analyze both simple 
and complex experimental designs involving multiple experimental conditions.

\section{Getting started}

\subsection{Installation}
ASpli is available at Bioconductor site and can be downloaded using
\texttt{biocLite()}:
<<installation, echo=TRUE, eval=FALSE>>=
source("https://www.bioconductor.org/biocLite.R")
biocLite("ASpli")
@
%<<loadASpli, echo=TRUE, eval=TRUE>>=
<<loadASpli, echo=TRUE, eval=FALSE>>=
library(ASpli)
@
 
\texttt{biocLite()} will take care of installing all the packages that ASpli 
depends on e.g. edgeR, GenomicFeatures, GenomicRanges, GenomicAlignments, Gviz, 
and other R package required. Some packages depend on operating system
packages, like \texttt{curl}, that are not installed automatically, and should
be installed by the user.

\subsection{Required data}

ASpli requires a single \texttt{BAM} file for each sample in the experiment.
Also, a gene annotation file is mandatory (usually a \texttt{GTF}/\texttt{GFF}
file). Make sure all files use the same coordinate system.

\subsection{Building a \texttt{TxDb} of your genome}
Gene annotation is handled using \texttt{TxDb} objects, which are provided by
\texttt{GenomicFeatures} package. The building of a \texttt{TxDb} object might 
be time consuming, depending on the size of your genome. If a \texttt{TxDb} has
been already created, we strongly recommend saving it as a \texttt{sqlite} file
and reloading it any time the analysis is run. The \texttt{GenomicFeatures} 
package provides functions to create TxDb objects based on data downloaded from
UCSC Genome Bioinformatics, BioMart or gff/gtf files. See "Creating New TxDb 
Objects or Packages" in the \texttt{GenomicFeatures} vignette for more details.
In this example, a TxDb is built from a gtf file:

<<makeTx, echo=TRUE, eval=FALSE>>=
library(GenomicFeatures)
TxDb <- makeTxDbFromGFF(
  file="genes.gtf",
  format="gtf")
@

\section{Running \texttt{ASpli}}

The workflow intended to be used with \texttt{ASpli} is divided in five
steps (Figure \ref{fig:ASpliStructure}). For each of these steps
ASpli provides a single or few functions to use:
\begin{enumerate}
  \item Reading of genomic alignments: \texttt{ loadBAM() }
  \item Feature extraction: \texttt{ binGenome() }
  \item Read counting: \texttt{ readCounts() }
  \item Alternative splicing discovery using junctions: \texttt{ AsDiscover() }
  \item DE and DU estimation \texttt{ DUreport(), junctionDUReport() and
  DUreportBinSplice() })
\end{enumerate}

At each step it is possible to export results in a friendly manner. 
See Section \ref{sec:output}:\nameref{sec:output} for more details.

ASpli defines four major objects classes used to store data and perform
operations related to a single task. Throughout this vignette each of these will
be discussed in detail, however they are presented here:
\begin{description}
  \item[ASpliFeatures] Contains data extracted from annotation, such genomic
  coordindates of genes.
  \item[ASpliCounts] Contains counts of aligned reads to each genomic feature.
  \item[ASpliDU] Contains results of Differential expression and usage analysis.
  \item[ASpliAS] Contains data useful to search novel splicing events.
\end{description}

\begin{figure}[ht!]
\centering
\includegraphics[width=0.85\textwidth]{images/workflow.pdf}
\caption{ Workflow of ASpli pipeline and prerequisites.  Each arrow represents
one of the main steps of a typical analysis. Rounded boxes are objects created
by ASpli functions.  Names under arrows are functions included in ASpli used to
generate each kind of object. }
\label{fig:ASpliStructure}
\end{figure}

\subsection{Binning the genome}
% Agregar definicion de los bins Io
\label{sec:binDefinition}
Sub-genic features such as exons and introns are analyzed using existing
annotations by splitting the genome into non-overlapping features called bins, 
following the methodology proposed for DEXSeq \cite{pmid22722343}. Exon and intron 
coordinates are then extracted from the annotation, but only those from 
multi-exonic genes are saved for further evaluation. When more than one isoform 
exists, some exons and introns will overlap. Exons and introns are then 
subdivided into new features called exon and intron bins, and are then 
classified into \emph{exclusively exonic bins}, \emph{exclusively intronic bins}, or 
\emph{alternative splicing bins} (AS) (See Figure \ref{fig:binDefinition}). 
In addition to these non overlapping bins, an additional type of bin is also
defined. The latter corresponds to full introns extracted from annotation,
before they are splitted by the presence of overlapping exons in another
isoform. These bins as referred as \emph{intron original} bins (Io).
% Sugerencia de Juli: explicar el por que de la clasificacion de los bins Io

\begin{figure}[ht!]
\centering
\includegraphics[width=12cm]{images/binDefinition.pdf}
\caption{ Schema of resulting bins from a gene with three hypothetical
  transcripts. Those bins that are exonic and intronic in different isoforms are 
  named \textit{AS bins}.
}
\label{fig:binDefinition}
\end{figure}

%Annotated junctions from all the transcripts are also reported. Junction 
%coordinates are defined as the last position of the five prime exon (donor 
%position) and first position of the three prime exon (acceptor).

\subsubsection{Splicing event assignation} \label{sec:eventAssign}
Each AS bin is assigned to a kind of splicing event. The assignation is
made using \emph{minimum  local gene models} considering unique patterns of three
bins: the bin being evaluated and the two neighbor bins (one to the left, one to 
the right). Figure (\ref{fig:binAssignation}) shows examples of  splicing event
categorization for different 
\emph{minimum local gene models}.

When the minimum local gene model contains two isoforms (see first row of figure
\ref{fig:binAssignation}) bins are labelled as follow:

\begin{itemize}
  \item \textbf {ES} Exon skipping
  \item \textbf {IR} Intron retention
  \item \textbf {ALt5'SS} Alternative 5' splicing site
  \item \textbf {ALt3'SS} Alternative 3' splicing site
\end{itemize}

When the minimum gene model contains three, four or five isoforms (no
more than five non redundant isoforms can exist in the minimum gene model),
labels of bins are similar to the ones defined previously but have an extra '*'
character ( \textbf{IR*}, \textbf{ES*}, \textbf{Alt5'SS*}, 
\textbf{Alt3'SS*}) in order to show that more than two isoforms exists and, 
therefore, the results should be used with caution (see rows 2-4 of figure 
\ref{fig:binAssignation}). 
This type of splicing events involve intronic subgenic regions sourrounded by 
exons in at least one isoform (\textbf{IR*}), and exonic subgenic regions 
sourrounded by two introns in at least one isoform (\textbf{ES*}). Minimum 
local gene models that do not fit in those categories are typified as 
\textbf{Alt5'SS*} or \textbf{Alt3'SS*} if they present exonic regions 
sourrounded by intronic and exonic neighbor subgenic regions.

In some cases, it is not possible to make a clear assignation for a bin with 
assumptions made here, those bins are labelled as \textbf{Undefined AS} (last
row of figure \ref{fig:binAssignation}).

Additionally, those bins that overlap with the beginning or ending of any transcript
are labelled as \textbf{external}. Please note that an external bin of a transcript
may overlap to a non external bin of another transcript, in these cases the bin
is still labelled as \textbf{external}. 


\begin{figure}[ht!]
\centering
\includegraphics[width=0.9\textwidth]{images/event_assignment.pdf}
\caption{ Summary of assignation of splicing events to bins from minimum gene
model. The bin being evaluated has a green background highlight. The blue boxes
represents exons, while the little light orange boxes represent introns. Gene
models shown are plus sense strand. }
\label{fig:binAssignation}
\end{figure}


\subsubsection{binGenome method }
\label{sectionBinG}

The \texttt{binGenome} method is a one-stop function to perform:
\begin{itemize}
 \item subgenic splitting of genes into bins
 \item junction, gene and bin coordinates extraction
 \item bin splicing event assignation
\end{itemize}

%\subsubsection{Naming conventions for bins and junctions }
%\label{sectionBinG}
Subgenic features are named as follow. For an hypothetical gene named GeneAAA:
\begin{itemize}
  \item \textbf{GeneAAA:E001}: defines first exonic bin
  \item \textbf{GeneAAA:I001}: defines first intronic bin
  \item \textbf{GeneAAA:Io001}: defines first \texttt{Intron original} before 
  being divided
  \item \textbf{GeneAAA:J001}: defines first annotated junction of GeneAAA
  \item \textbf{chr.start.end}: defines an experimental junction, \texttt{chr}
  is the chromosome name, \texttt{start} is the leftmost position of
  the junction and \texttt{end} is the rightmost position of the junction.   
\end{itemize}

Bins and junctions are consecutively named from 5' to 3' sense of the plus
strand of the reference sequence. This implies that  bins and junctions with
lower numbering are always closer to the 5' of that strand. Alternative splicing
bins are named as exons.

%All tasks presented in this section: subgenic splitting of genes into
%bins, coordinates extraction of junctions, genes and bins, naming of genes and
%bins and splicing event assignation to bins are performed by single method,
%\texttt{binGenome}. The result is an object of class \texttt{ASpliFeatures}.
%Method \texttt{binGenome} accepts an additional optional argument
%\texttt{geneSymbols} used to assign symbols to genes (i.e. common name or a short
%description).
%Then, features coordinates could be obtained using their accesors methods.
%Method \texttt{featuresg} returns coordinates for genes as a GRangesList
%object, and methods \texttt{featuresb} and \texttt{featuresj} returns
%coordinates for bins and bins respectively as a GRanges object. GRangesList and
%GRanges are objects defined in GenomicRanges package. These objects contains
%aditional information stored as metadata and can be accessed with \texttt{mcols}
%function.


\texttt{binGenome}'s output is an object of class \texttt{ASpliFeatures}.

Methods \texttt{featuresg}, \texttt{featuresb}, \texttt{featuresj} can be used to
access genes, bins and junctions coordinates as \texttt{GRanges-List} (for the first case)
or as \texttt{GRanges} objects (for bins and junctions).
\texttt{GRangesList} and \texttt{GRanges} are objects defined in the \texttt{GenomicRanges} package. These objects contains
aditional information stored as metadata that can be accessed with \texttt{mcols} function.

<<binGenome, echo=TRUE, eval=FALSE>>=
library(GenomicFeatures)
annFile       <- aspliExampleGTF()
aTxDb         <- makeTxDbFromGFF(annFile)
# extract features from annotation
features      <- binGenome( aTxDb ) 
# Accesors of ASpliFeatures object 
geneCoord     <- featuresg( features )
binCoord      <- featuresb( features )
junctionCoord <- featuresj( features )
# Extract metadata annotation, such as bins names, event and feature type
binMetadata   <- mcols( binCoord )
@

<<binGenome, echo=TRUE, eval=FALSE>>=
# extract features and map symbols to genes
symbols       <- data.frame( row.names = genes( aTxDb ), 
                             symbol = paste( 'This is symbol of gene:',
                                             genes( aTxDb ) ) )
features      <- binGenome( aTxDb, geneSymbols = symbols ) 
features
@

\subsection{Defining the targets}
\label{sec:targetsDef}

Sample names, genomic alignment file names and experimental factors should be specified in a 
\texttt{data.frame} that has as many rows as samples. Each row should be
named with the corresponding sample name. The first column must be named \texttt{bam} and
should contain the path to a bam file. The remaining columns should be used to state the corresponding 
experimental factors for each sample. This 
\texttt{data.frame} is referred to as the \textbf{targets} \texttt{data.frame}. 

For example, for a one factor experiment with two conditions (Control and
Mutant) and two replicates for each condition, the \textbf{targets}
\texttt{data.frame} could be defined as follow:
<<targetsDF, echo=TRUE, eval=TRUE>>=
bamFiles <- c( "Ct1.bam", "Ct2.bam", "Mut1.bam","Mut2.bam" )
targets <- data.frame( row.names =  c("CT_rep1","CT_rep2", "Mut_rep1", "Mut_rep2"),
                       bam = bamFiles,
                       genotype = c("CT","CT", "Mut", "Mut") , 
                       stringsAsFactors = FALSE )
targets
@ 

More sofisticated designs should be specified as follow. For example: for a two
factor experiment (factors genotype and time), with two values for each factor
adn two replicates for each condition, the \textbf{targets} \texttt{data.frame}
can be defined as follow:
<<targetsDF2, echo=TRUE, eval=TRUE>>=
bamFiles <- c("Ct_time1_rep1.bam", "Ct_time1_rep2.bam",
              "Ct_time2_rep1.bam", "Ct_time2_rep2.bam",
              "Mut_time1_rep1.bam","Mut_time1_rep2.bam",
              "Mut_time2_rep1.bam","Mut_time2_rep2.bam")
targets_2 <- data.frame( row.names = c( 'CT_t1_r1',  'CT_t1_r2',
                                        'CT_t2_r1',  'CT_t2_r2',
                                        'Mut_t1_r1', 'Mut_t1_r2',
                                        'Mut_t2_r1', 'Mut_t2_r2' ),
                         bam = bamFiles,
                         genotype = c( 'CT',  'CT',  'CT',  'CT', 
                                       'MUT', 'MUT', 'MUT', 'MUT' ),
                         time     = c( 't1', 't1', 't2', 't2', 
                                       't1', 't1', 't2', 't2' ),
                         stringsAsFactors = FALSE )
targets_2
@

The name of the each experimental condition is determined from the experimental
factors defined in the \textbf{targets} \texttt{data.frame}. They can be
consulted using the \texttt{getConditions} function. 

<<targetsDF2, echo=TRUE, eval=FALSE>>=
# Show the name of each unique condition in the console
getConditions( targets_2 )
@

\subsection{Read Counting}

\subsubsection{Genomic alignment files loading} 
\label{secGAFL}
Once the \textbf{targets} \texttt{data.frame} is defined the alignment data
from bam files can be loaded easily with \texttt{loadBAM} function. BAM files
are loaded and stored as a named list of \texttt{GAlignments} objects.
\texttt{GAlignments} objects are defined in \texttt{GenomicAlignments} package.

<<loadBam, echo=TRUE, eval=FALSE>>=
targets <- aspliTargetsExample()
bam <- loadBAM(targets)
@

%\subsubsection{Counting alignment reads overlapping features } 
%The method \texttt{readCounts} allows to count the number of reads from genomic
%alignments that overlaps each feature of interest. These features are genes,
%bins, junctions, and intron/exon flanking regions.
%For genes and bins, a read densitie value is calculated as shown below:
%\begin{equation*}{
%  f_{d} = \frac{ N_{r} }{ l }
%}
%\end{equation*}
%Where $f_{d}$ is the feature read density, $N_{r}$ is the number of reads
%overlaping the feature and $l$ is the length of the feature.
%
%Results generated by \texttt{readCounts} are stored into an \texttt{ASpliCounts}
%object. The usage of this method is shown below:
%
%<<readCounts, echo=TRUE, eval=FALSE>>=
%counts <- readCounts ( 
%    features, 
%    bam, 
%    targets, 
%    cores = 1, 
%    readLength = 100L, 
%    maxISize = 50000,
%    minAnchor = NULL )
%@
%Where the accepted parameters are:


\subsubsection{readCounts: Counting feature overlapping reads} 
\label{sec:rcounts}
The method \texttt{readCounts} allows to count the number of reads that overlaps each feature of interest (i.e. genes,
bins, junctions, and intron/exon flanking regions).
For genes and bins, read density values are also calculated considering the ratio 
between the number of overlapping reads and the length of a given feature.
The following parameters can be specify:

\begin{itemize}
  \item \texttt{features}: An ASpliFeatures object. 
  \item \texttt{bam}: A list of \texttt{GAlignments} generated with loadBAM
\texttt{function}.
  \item \texttt{targets}: A \textbf{targets} \texttt{data.frame} as defined in
section \secref{sec:targetsDef}. 
  \item \texttt{cores}: Number of proccessor cores used to speed up the
computation.
  \item \texttt{readLength}: Read length of sequenced library 
  \item \texttt{maxISize}: maximum intron expected size. Junctions longer than
this size will be dicarded \cite{Hong01122006}.
  \item \texttt{minAnchor}: minimum anchoring overlap for intron flanking
  regions (see \secref{sec:intronFlanking} ). 
\end{itemize}

<<readCounts, echo=TRUE, eval=FALSE>>=
counts <- readCounts ( 
    features, 
    bam, 
    targets, 
    cores = 1, 
    readLength = 100L, 
    maxISize = 50000,
    minAnchor = NULL )
@


The result of \texttt{readCounts} method is an object of class 
\texttt{ASpliCounts}. Count and read density data could be extracted from it as
\texttt{data.frame} objects using accesors methods. Also, it is possible to 
export count and read densities tables to text files in a tidy manner.

<<readCountAccessors, echo=TRUE, eval=FALSE>>=
# Accessing count and read density data
GeneCounts <- countsg(counts)
GeneRd <- rdsg(counts)

BinCounts <- countsb(counts)
BinRd <- rdsb(counts)

JunctionCounts <- countsj(counts)

# Export count data to text files 
writeCounts(counts=counts, output.dir = "example")
writeRds(counts=counts, output.dir = "example")
@

\subsubsection{Counting intron flanking regions }
\label{sec:intronFlanking}

Every intron is considered as a potential retained intron. 
In order to analyze putative retention events for intron I, \texttt{ASpli} considers the 
corresponding adjacent 5' and 3' exons (E1 and E2, following
the order in the genomic coordinates, regardless of the sense of the strand). Then, following 
\cite{pmid25258385}, new artificial ranges that overlap the two retention regions E1I 
(connecting exon E1 and intron I) and IE2 (connecting intron I and exon E2) are defined:

\begin{itemize}
  \item E1I: intron start - (readLength - minAnchor) --- intron start + (
  readLength - minAnchor )
  \item IE2: intron end - (readLength - minAnchor) --- intron end + ( readLength
  - minAnchor )
  \item \texttt{ minAnchor } is 10\% of read length by default (parameter
  \texttt{minAnchor} )
\end{itemize}

% TODO - EXPLICAR MEJOR  \\
Only those reads which minimum overlap \textit{readLength} and without gap in 
this interval are considered. Accordingly, only sequences aligned within those 
two exons/introns are counted. If the reads of your genomic alignment are
trimmed by quality rendering reads with a length inferior to the nominal
read length, then \texttt{E1I} and \texttt{IE2} values will be underestimated.


<<readCountAccessors2, echo=TRUE, eval=FALSE>>=
# Accessing counts for intron flanking regions
e1iCounts <- countse1i(counts)
ie2Counts <- countsie2(counts)
@

\subsubsection{Additional considerations}

For a given gene, the count number is computed as
the number of reads overlapping any exon included in the corresponding annotated gene model. If a
single read overlaps more than one exon, it is counted only once. Note that a read can
overlap two different genes, in this case it is counted for both of them.
 
In order to estimate read densities, effective gene lengths are
considered. The effective length is calculated as the sum of the length of
exonic bins and alternative bins (i.e. all bins except intronic bins).

%Some reads might overlap to more than one bin and could be counted several
% times.
% Hay que aclarar en que situaciones ocurre esto.

Junctions are extracted from BAM files. They are defined as those reads
that aligned against disjoint region of the reference genome(N operator of CIGAR notation
for aligned reads \cite{pmid19505943} ), and are essential piece of information for alternative
splicing event quantification and discovery. Junction alignment
quality/confidence is extremely important and it should be controlled at the
moment of the alignment step.

\subsubsection{The ASpliCounts object }
\label{sec:countsContents}
%Accessor methods of \texttt{ASpliCounts} object return \texttt{data.frame}
%objects for genes counts, genes read density, bin counts, bin read density, E1I
%flanking region counts and IE2 flanking region counts. All of them share the
%same structure, however the specific content is different for each kind of 
%feature. The first columns contains feature information extracted from the
%annotation, and then there several columns with count or read density data
%for each experimental sample. A more detailed overview is shown below:

Gene, bin and junction information stored in an \texttt{ASpliCounts} object can 
be accessed  as stated in section (\ref{sec:rcounts}) and (\ref{sec:intronFlanking}).
Accesor functions returns \texttt{data.frame} objects with the required information.
They typically share the same general structure, but also include specific content depending on the
nature of the requested feature.  A more detailed overview is shown below:

\begin{itemize}
  \item Gene counts (\texttt{countsg}) and gene read densities (\texttt{rdsg})
  \texttt{data.frames} contain the following information (see table 
  \ref{tab:counts.gene} for an example):
    \begin{description}
      \item[row.names] Gene name as reported in annotation data.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section
      \secref{sec:binDefinition}).
      \item[locus\_overlap] Names of overlapping \textit{loci}.
      \item[gene\_coordinates] Genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[start] Left-most position of the gene.
      \item[end] Right-most position of the gene.
      \item[length] Number of bases covering the gene.
      \item[effective\_length] Number of bases that overlaps with annotated exons.
      \item[sample data] Gene counts or read density (one column per sample).
      \\
    \end{description} 

  \item Bin counts (\texttt{countsb}) and bin read densities (\texttt{rdsb}) \texttt{data.frames} contain the
  following information:
    \begin{description}
      \item[row.names] Bin name assigned according to the convention presented in \secref{sectionBinG}.
      \item[feature] Type of bin: \textbf{E} for exonic bins, \textbf{I}
      for intronic bins and \textbf{Io} for introns before splitting.
      \item[event] Splicing event asigned to the bin (see section
      \ref{sec:binDefinition}:\nameref{sec:eventAssign})
      \item[locus] The name of the locus that contains the bin.
      \item[locus\_overlap] Names of all other overlapping \textit{loci}.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section \secref{sec:binDefinition}).
      \item[gene\_coordinates] Genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[start] Left-most position of the bin.
      \item[end] Right-most position of the bin.
      \item[length] Number of bases covering the bin.
      \item[sample data] Bin counts or read density (one column per sample)
      densities.
      \\
    \end{description}

  \item Junction counts (\texttt{countsj}) \texttt{data.frame} contains the
  following information:
    \begin{description}
      \item[row.names] Junction name in format
      \texttt{chromosome.start.end} (see \secref{sectionBinG}).
      \item[junction] If junction coincides with a junction inferred
      from the annotation, the name is shown as is given in section
      \ref{sectionBinG}:\nameref{sectionBinG}, otherwise it contains \texttt{noHit}.
      \item[gene] Locus that contains the junction.
      \item[strand] Sense strand of the gene.
      \item[multipleHit] \texttt{yes} if junction spans multiple genes.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section
      \secref{sec:binDefinition}).
      \item[gene\_coordinates] Genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[bin\_spanned] Names of the bins spanned by the junction.
      \item[j\_within\_bin] If junction falls within a single bin, the name of
      that bin is shown.
      \item[sample data] Junction counts (one column per sample).
      \\
    \end{description}

  \item E1I (\texttt{e1icounts}) and IE2 (\texttt{ie2counts}) flanking region count \texttt{data.frame} contain the
  following information:
    \begin{description}
      \item[row.names] Junction name in format \texttt{chromosome.start.end}
      (see \secref{sectionBinG}).
      \item[junction] If junction coincides with a junction inferred
      from the annotation, the name is shown as is given in section
      \ref{sec:binDefinition}:\nameref{sec:binDefinition}, otherwise
      contains \texttt{noHit}.
      \item[gene] Name of the locus that contains the junction.
      \item[strand] Strand sense of the gene.
      \item[multipleHit] \texttt{yes} if junction spans multiple
      genes.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section \ref{sec:binDefinition}:\nameref{sec:binDefinition}).
      \item[gene\_coordinates] Show the genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[bin\_spanned] Names of the bins spanned by the junction.
      \item[j\_within\_bin] If junction falls within a single bin, the name of
      that bin is shown.
      \item[sample data] Junction counts (one column per sample).
      \\
    \end{description}
    
\end{itemize}

\begin{table}
\centering
\begin{tabular}{lllllllllll}
\rotatebox{90}{Row names} & 
\rotatebox{90}{symbol} &
\rotatebox{90}{locus\_overlap} &
\rotatebox{90}{gene\_coordinates} &
\rotatebox{90}{start} &
\rotatebox{90}{end} &
\rotatebox{90}{length} &
\rotatebox{90}{effective\_length} &
\rotatebox{90}{Sample 1} & 
\rotatebox{90}{Sample 2} &
\rotatebox{90}{ ... } \\ \hline
GENE01 & GENE01 & - & chr1:1-700     & 1    &  700 & 700 & 700 & 324 & 314 & n\\
GENE02 & GENE02 & - & chr1:1001-1800 & 1001 & 1800 & 800 & 550 & 327 & 333 & n\\
GENE03 & GENE03 & - & chr1:2001-2800 & 2001 & 2800 & 800 & 650 & 342 & 321 & n\\  
GENE04 & GENE04 & - & chr1:3001-3800 & 3001 & 3800 & 800 & 650 & 313 & 337 & n\\  
...    & ...    &...& ...            & ...  & ...  & ... & ... & ... & ... &
...
\end{tabular}
\caption{Gene counts table example.}
\label{tab:counts.gene}
\end{table}
%
%\begin{table}
%\centering
%\begin{tabular}{lllllllllllll}
%\rotatebox{90}{Row names} & 
%\rotatebox{90}{feature} &
%\rotatebox{90}{event} &
%\rotatebox{90}{locus} &
%\rotatebox{90}{locus\_overlap} &
%\rotatebox{90}{symbol} &
%\rotatebox{90}{gene\_coordinates} &
%\rotatebox{90}{start} &
%\rotatebox{90}{end} &
%\rotatebox{90}{length} &
%\rotatebox{90}{Sample 1} & 
%\rotatebox{90}{Sample 2} &
%\rotatebox{90}{ ... } \\ \hline
%GENE01:E001 & E & external & GENE01 & - & GENE01 & chr1:1-700     &
%1    & 300  & 300 & 189 & 199 & n \\ 
%GENE01:E002 & E & IR       & GENE01 & - & GENE01 & chr1:1-700     &
%301  & 500  & 200 & 154 & 143 & n \\
%GENE01:E003 & E & external & GENE01 & - & GENE01 & chr1:1-700     &
%501  & 700  & 200 & 129 & 131 & n \\
%GENE02:E001 & E & external & GENE02 & - & GENE02 & chr1:1001-1800 & 
%1001 & 1250 & 250 & 197 & 206 & n \\
%...         &   & ...      & ...    &...& ...    & ...            &
%...  & ...  & ... & ... & ... & ... 
%\end{tabular}
%\caption{Bins counts table example.}
%\label{tab:counts.bins}
%\end{table}
%
%
%\begin{table}
%\centering
%\footnotesize 
%\begin{tabular}{ l l l l l l l p{3cm} l l l l}
%\rotatebox{90}{Row names} & 
%\rotatebox{90}{junction} &
%\rotatebox{90}{gene} &
%\rotatebox{90}{strand} &
%\rotatebox{90}{multipleHit} &
%\rotatebox{90}{symbol} &
%\rotatebox{90}{gene\_coordinates} &
%\rotatebox{90}{bin\_spanned} &
%\rotatebox{90}{j\_within\_bin} &
%\rotatebox{90}{Sample 1} & 
%\rotatebox{90}{Sample 2} &
%\rotatebox{90}{ ... } \\ \hline
%chr1.300.501   & GENE01:J001 & GENE01 & + & - & GENE01 & chr1:1-700     & 
%GENE01:E001; GENE01:E002; GENE01:E003 & - & 30 & 24 & n \\
%chr1.1250.1301 & GENE02:J001 & GENE02 & + & - & GENE02 & chr1:1001-1800 & 
%GENE02:E001; GENE02:E002              & - & 60 & 64 & n \\
%chr1.1250.1601 & GENE02:J003 & GENE02 & + & - & GENE02 & chr1:1001-1800 & 
%GENE02:E001; GENE02:E002; GENE02:E003 & - & 27 & 26 & n \\
%chr1.1400.1601 & GENE02:J002 & GENE02 & + & - & GENE02 & chr1:1001-1800 & 
%GENE02:E002; GENE02:E003              & - & 47 & 49 & n \\
%...            &             & ...    &...&...& ...    & ...            & 
%...  & ...  & ... & ... & ... 
%\end{tabular}
%\caption{Junctions counts table example.}
%\label{tab:counts.junctions}
%\end{table}
%
%\begin{table}
%\centering
%\begin{tabular}{llllllllllll}
%\rotatebox{90}{Row names} & 
%\rotatebox{90}{event} &
%\rotatebox{90}{locus} &
%\rotatebox{90}{locus\_overlap} &
%\rotatebox{90}{symbol} &
%\rotatebox{90}{gene\_coordinates} &
%\rotatebox{90}{start} &
%\rotatebox{90}{end} &
%\rotatebox{90}{length} &
%\rotatebox{90}{Sample 1} & 
%\rotatebox{90}{Sample 2} &
%\rotatebox{90}{ ... } \\ \hline
%GENE02:I001  & -   & GENE02 & - & GENE02 & reference:1001-1800 & 1161 &  1341 & 181 & 0  & 0  & n \\
%GENE03:Io002 & -   & GENE03 & - & GENE03 & reference:2001-2800 & 2161 &  2341 & 181 & 7  & 6  & n \\
%GENE05:Io001 & -   & GENE05 & - & GENE05 & reference:4001-4800 & 4161 &  4341 & 181 & 31 & 26 & n \\
%GENE07:E002  & IR* & GENE07 & - & GENE07 & reference:6001-6800 & 6161 &  6341 & 181 & 20 & 23 & n \\
%...          &     & ...    &...& ...    & ...                 & ...  & ...   &
%...          & ... & ...    &...
%\end{tabular}
%\caption{Exon/Intron flanking regions counts table example.}
%\label{tab:counts.e1i}
%\end{table}
%


\subsection{AsDiscover: A Junction-centered analysis of alternative splicing}

Count data of bins and junctions can be used to get an integrative view of 
alternative splicing events. The method \texttt{AsDiscover} will produce an
\texttt{ASpliAS} object containing comprehensive results that can be used as
supporting evidence of alternative splicing usage of bins. 

<<AsDiscover, echo=TRUE, eval=FALSE>>=
as <- AsDiscover( counts, targets, features, bam, readLength=100L, 
                  threshold = 5)
@
where:
\begin{itemize}
  \item \texttt{counts}: An ASpliCount object.
  \item \texttt{targets}: A \textbf{targets} \texttt{data.frame} as defined in
section \secref{sec:targetsDef}.  
  \item \texttt{features}: An ASpliFeatures object (see \secref{sectionBinG}). 
  \item \texttt{bam}: A list of \texttt{GAlignments} generated with loadBAM
\texttt{function} (see \secref{secGAFL}).
  \item \texttt{readLength}: Read length of sequenced library 
  \item \texttt{threshold}: minimum number of reads supporting a junction.
  \item \texttt{cores}: Number of proccessor cores used to speed up the
computation.
\end{itemize}

The analysis will consider junctions that are completely included within a
unique gene and have more than a minimum number of reads supporting them, by
default this number is five.

For each experimental junction identified, it is also reported if it is novel
and which bins are spanned. In addition, it is stated if the junction is 
completely included in an annotated bin, which would indicate that the AS event
is a possible \emph{exintron} \cite{pmid25934563}.


The contents of an \texttt{ASpliAS} object can be retrieved using accessor
methods: \texttt{irPIR}, \texttt{altPSI}, \texttt{esPSI},
\texttt{joint} \texttt{\mbox{junctionPIR}}, \texttt{\mbox{junctionPSI}} (see
section \secref{secASpliAsContent} for a detailed description of each method and 
retrieved data). In addition, method \texttt{writeAS}, allows to easily export 
all the data into text files tables.

<<writeAS, echo=TRUE, eval=FALSE>>=
# Access result from an ASpliAS object 
irPIR  <- irPIR( as )
altPSI <- altPSI( as )
esPSI  <- esPSI( as )
junctionsPIR <- junctionsPIR( as )
junctionsPSI <- junctionsPSI( as )
allBins      <- joint( as )
# Export results to files
writeAS(as=as, output.dir="example")
@

\subsubsection{PIR and PSI metrics}
\label{sec:psir}
To provide an integrative view of the AS events being analyzed, splice junction
information is used to analyze differential bin usage. PSI (percent of
inclusion) and PIR (percent of intron retention) metrics, which are extensively
used to quantify AS \cite{pmid21057496}, are calculated for bins junctions (see
figure \ref{fig:pirEq} ). 
Both, the PIR and PSI metrics, strongly enrich the count-centric analysis and 
provide independent experimental support for the existence of novel AS events, 
based on the identification of corresponding changes in nearby splice junctions.

Only junctions that pass an abundance filter are considered for PSI/PIR 
calculation: a minimum number of counts (given by the threshold parameter) 
should be attained in all samples of at least one condition in order to
consider the corresponding junction for further analysis (the default value 
of this threshold is five counts).


%\begin{figure}[ht!]
%    \begin{subfigure}[t]{1\textwidth}
%    \centering
%      \includegraphics[width=0.75\textwidth]{images/psi5ss.pdf}
%      \includegraphics[width=0.75\textwidth]{images/psi3ss.pdf}      
%      \caption{PSI metric for alternative 5' and 3' splicing sites estimation}
%    \end{subfigure}
%    \begin{subfigure}[t]{1\textwidth}
%    \centering
%      \includegraphics[width=0.75\textwidth]{images/psies.pdf}
%      \caption{PSI metric for exon skipping estimation}
%    \end{subfigure}
%    \begin{subfigure}[t]{1\textwidth}
%      \centering
%      \includegraphics[width=0.75\textwidth]{images/pir.pdf}
%      \caption{PIR metric for intron retention estimation}
%    \end{subfigure}    
%    \caption{PSI and PIR metrics estimation and their relationship with
%    junctions}
%    \label{fig:pirEq}
%\end{figure}

\begin{figure}[ht!]
    \centering
      \includegraphics[width=0.75\textwidth]{images/psi_pir2.pdf}
    \caption{PSI and PIR metrics estimation and their relationship with
    junctions}
    \label{fig:pirEq}
\end{figure}


For each bin, a PIR or a PSI  metric is calculated, depending on the splicing
event category assigned to that bin (see section \secref{sec:eventAssign}). If
no splice event was assigned, meaning that the bin is not alternative, an exon
will be considered to be involved in a putative exon skiping splicing event, and
an intron will be considered to be involved in a putative intron retention 
splicing event.

Every detected junction implies the existence of an intron that could be potentially 
retained. Therefore a PIR value is computed for every junction. Also, every
junction defines a 5' and a 3' splicing site. Each one of these can be found
exclusively in this junction or found also in other junctions. The latter is
evidence of an exon skipping event or an alternative 5' or 3' splicing site
event. Althought is not possible to infer the kind of event, the computation of
a PSI value is helpful to detect this cases. The details of the formulas used to
calculte these values are shown in figure \ref{fig:psir_junc}.

%This information allows the user to obtain reliable information on the relative
%abundance of the AS event being evaluated. Both metrics strongly enrich the 
%count-centric analysis and provide independent experimental support for the 
%existence of novel AS events, based on the identification of corresponding 
%changes in nearby splice junctions.

% TODO - PONER EN NEGRITA EL TIPO DE EVENTO (EN LA COLUMNA IZQUIERDA)
%\begin{figure}[ht!]
%  \begin{subfigure}[t]{1\textwidth}
%    \centering
%    \includegraphics[width=0.75\textwidth]{images/pir_junc.pdf}
%    \caption{ PIR metric to estimate intron retention for junctions}
%  \end{subfigure}
%  \begin{subfigure}[t]{1\textwidth}
%    \centering
%    \includegraphics[width=0.75\textwidth]{images/psi_junc.pdf}
%    \caption{ PSI metric to compute inclusion of exons bearing a junction.} 
%  \end{subfigure}
%  \caption{PSI and PIR metrics for junctions}
%  \label{fig:psir_junc}
%\end{figure}

\begin{figure}[ht!]
  \centering
  \includegraphics[width=0.75\textwidth]{images/psi_pir_junc.pdf}
  \caption{PSI and PIR metrics for junctions}
  \label{fig:psir_junc}
\end{figure}

\subsubsection{New splicing events discovery} 
\texttt{ASpli} allows novel AS events to be identified based on the splice
junction repertoire. A novel AS event is identified whenever a
novel splice junction that shares its beginning or end with another splice 
junction is discovered. When a novel AS event is identified using the splice
junction repertoire, the PSI metric is calculated and reported. This ability to
detect novel AS events and to estimate the magnitude of the changes in the usage
of these AS events under different experimental conditions are original
functions of the package.

\subsubsection{Contents of ASpliAS object}
\label{secASpliAsContent}

Accessor methods of \texttt{ ASpliAS } object return \texttt{data.frame}
objects containing detailed information for bins involved in intron retention,
5' or 3' alternative splicing, exon skipping events, and for experimental junctions. 

\begin{itemize}
  \item Results for bins involved in intron retention, alternative 5' or 3'
  splicing, and exon skipping events can be accessed with \texttt{irPIR},
  \texttt{altPSI} and \texttt{esPSI} respectively, or using \texttt{joint}
  method for all of them altogether. These methods return \texttt{data.frame}
  objects which report inclusion and/or exclusion evidence considering counts assigned to
  \textbf{J1}, \textbf{J2}, and \textbf{J3} type of junctions (see figures \ref{fig:pirEq}
  and \ref{fig:psir_junc}):\\


    \begin{description}
      \item[row.names] Bin name as is defined in \ref{sectionBinG}.
      \item[event] Splicing event asigned to the bin (see section
      \ref{sec:binDefinition}:\nameref{sec:eventAssign})
      \item[J1] The name of J1 junction or E/I flanking region.
      \item[sample data for J1] J1 counts for each sample.
      \item[J2] The name of J2 junction or E/I flanking region.
      \item[sample data for J2] J2 counts for each sample.
      \item[J3] The name of J3 junction or E/I flanking region.
      \item[sample data for J3] J3 counts for each sample.
      \item[PSI or PIR data] PSI variant metric or PIR metric (one column per sample).
      \\
    \end{description}
  \item Results for experimental junctions assuming that they correspond to
  introns involved in intron retention events can be accessed with
  \texttt{junctionPIR} method. The returning value is a \texttt{data.frame} with
  columns:
    \begin{description}
      \item[row.names] Junction name in format \texttt{chromosome.5'site-3'site}.
      \item[hitIntron] If it exists, the name of a bin that exactly match the positions of
      the junction.
      \item[hitIntronEvent] If \texttt{hitIntron} is not empty, the
      splicing event assigned to that bin.
      \item[event] Splicing event asigned to the bin (see section
      \ref{sec:binDefinition}:\nameref{sec:eventAssign})
      \item[sample data for J1] J1 counts for each sample.
      \item[sample data for J2] J2 counts for each sample.
      \item[sample data for J3] J3 counts for each sample.
      \item[PIR data]  PIR metric (one column per sample).
      \\
    \end{description}
  \item Results for experimental junctions assuming that the bearing
  exons are involved in alternative splice events can be accessed with
  \texttt{junctionPSI} method. The returning value is a \texttt{data.frame} with
  columns:
    \begin{description}
      \item[row.names] Junction name in format \texttt{chromosome.5'site-3'site}.
      \item[junction] If junction coincides with a junction inferred
      from the annotation, the name is shown as is given in section
      \ref{sec:binDefinition}:\nameref{sec:binDefinition}, otherwise
      contains \texttt{noHit}.
      \item[gene] The name of the locus that contains the junction.
      \item[strand] Sense strand of the gene.
      \item[multipleHit] \texttt{yes} if junction spans multiple
      genes.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section
      \secref{sec:binDefinition}).
      \item[gene\_coordinates] Show the genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[bin\_spanned] The names of the bins spanned by the junction.
      \item[j\_within\_bin] If junction fall within a single bin, the name of
      that bin is shown.
      \item[sample data for J3] J3 counts for each sample.
      \item[StartHit] The name of J1 junction.
      \item[sample data for J1] J1 counts for each sample.
      \item[PSI data] A column by condition with the PSI metric for J1. The name
      of the column is the name of the condition with '.start' suffix.
      \item[EndHit] The name of J2 junction.
      \item[sample data for J2] J2 counts for each sample.
      \item[PSI data] A column by condition with the PSI metric for J2. The name
      of the column is the name of the condition with '.end' suffix.
      \\
    \end{description}  

\end{itemize}

\subsection{Estimating differential gene expression and bin/junction usage}

\texttt{ASpli} provides two methodolgies (\texttt{DUreport} and \texttt{DUreportBinSplice}) to perform statistical analysis of differential gene expression and differential usage of bins. Both functions implements the same gene-level analysis. However, while \texttt{DUreportBinSplice} is a wrapper to the \texttt{diffSpliceDGE} function 
from edgeR, \texttt{DUreport} implements a pseudocount normalization method that provides similar detetection performance. \texttt{ASpli} also provides a third function, \text{junctionDUreport}, that only takes into account junction information.

Results are stored in \texttt{ASpliDU} objects
and can be accessed for bins, junctions, and genes using the accesors
methods \texttt{binsDU}, \texttt{junctionDU}, and \texttt{genesDE}
respectively. Results can also be exported as text files tables using the \texttt{writeDU}
function.

\subsubsection{Differential gene expression (DE)}
\label{sec:de}

The first step of the gene DE analysis pipeline is the selection of those
annotated genes that are considered to be expressed, or at least that can be
confidently detected. For this, the average number of reads for each gene
and condition and the average read density for each gene and condition are
calculated as shown below:

\begin{eqnarray*}
   \hat G^{c}_{ij} = \sum_{k \in K_j} \frac{ S^{c}_{ik} }{ |K_j| } \\
   \hat G^{rd}_{ij} = \sum_{k \in K_j} \frac{ S^{rd}_{ik} }{ |K_j| }
   \label{eq:genByCond}
\end{eqnarray*}

Where $\hat G^{c}_{ij}$ is avg. gene count for gene $i$ in condition $j$,
$\hat G^{rd}_{ij}$ is avg. read density for gene $i$ in condition $j$, $S^{c}_{ik}$
is gene count for gene $i$ in sample $k$, $S^{rd}_{ik}$ is read density for
gene $i$ in sample $k$. $K_j$ is the set of samples corresponding to condition $j$.

Gene $i$ is considered expressed if any of the corresponding $\hat G^{c}_{ij}$
and $\hat G^{rd}_{ij}$ values are equal or greater than a minimum cutoff (by 
default, these numbers are 10 for gene count and 0.05 for read density) and for
any condition $j$.

Differential gene expression is estimated with edgeR \cite{Robinson2012} 
package using an exact test (for paired comparison) or generalized linear model
(GLM) for more complex designs. For each gene \textbf{logFC}, \textbf{pvalue},
and \textbf{adjusted  pvalue} by false discovery rate \cite{fdr} are reported. 

\texttt{DUreport} and \texttt{DUreportBinSplice} methods perform genes DE
analysis with the same methodology and gives the same results. 

Each condition in the analysis is weighted by contrast vector that ultimately
defines the comparison that will be done. For a simple pair comparison between
\texttt{Control} and \texttt{Trat} conditions, represeted as $Trat - Control$ 
can be expressed as:

<<export, echo=TRUE, eval=FALSE>>=
contrast <- c( -1, 1 ) 
@

Assuming that the order of the conditions given by \texttt{getConditions}
methods is \texttt{ [ 'Control', 'Trat' ]}. 

%The statistical computation is delegated in \texttt{egdeR} package capabilities.
%It offers an exact test to compute p-values when comparing two conditions, and
%fitting a generalized linear model for designs involving more conditions. ASpli
%selects the appropiate methodoly for the analysis in a transparent way without
%requiring the user to know the details.


\subsubsection{Differential bin and junction usage (DU)}
\label{sec:du}

Bins and junctions are treated similarly for DU analysis. 
In order to select informative events, only bins and junctions from expressed
genes (see section \secref{sec:de}) will be taken into account if they have more
reads than a threshold (five reads by default) supporting them in at least one 
of the conditions. 

External bins do not participate in alternative splicing, therefore are usually
excluded from the analysis. However, an external bin in one isoform can overlap
to a non external bin in other isoform that can participate in alternative
splicing regulation. ASpli allows to optionally include the external bins in
the analysis. Also, original intron bins can be optionally excluded. The
inclusion of those bins can affect the estimation of corrected p-values (fdr).
The information provided by intron original bins are highly correlated
with the one of their sub bins, and increase largely the number of
events to be analyzed. Therefore the fdr correction is more strict. Also, there
is a violation on the fdr correction assupmtion that all individual tests are
independent from each other. If a intron original bin show a significant change,
there is a very high chance that at least one of their sub bins also shows a
significant change.

Differential bin/junction usage are estimated using the statistical methodology
proposed in the \texttt{edgeR} package. Read counts from bins/junctions can not
be used naively for differential usage analysis. Counts needs to be adjusted
to remove the signal from changes in gene expression. To perform this, Aspli 
provides three alternatives: 

\begin{enumerate}
   \item Adjust bin/junction counts to gene counts
   \item Use an offset value given by gene expression
   \item Delegate differential bin usage estimation to \texttt{edgeR}'s 
         \texttt{ diffSpliceDGE }. This is available only for bins.
\end{enumerate}

The adjustment of bin counts to gene counts is performed by transforming the
raw bin counts. The transformation is done by:

\begin{equation}
  B^{A}_{ijk} = \frac{ B_{ijk} \times \bar{G_{j}} } { G_{jk} }
\end{equation}

Where $B^{A}_{ijk}$ is the adjusted bin count for bin $i$ of gene $j$ in sample
$k$, $B_{ijk}$ is the raw bin count for bin $i$ of gene $j$ in sample $k$, 
$\bar{G_{j}}$ is the mean raw count for gene $j$ for all samples and $G_{jk}$ is
the raw gene count for gene $j$ in sample $k$.

Then the adjusted bin counts are used as input for an statical analysis
identical to the one performed on genes using the methods provided by edgeR
package.

The offset methodology consist in define a vector that reflects the change of
gene expression levels and use it to correct the observed change in bin usage
in the statistical analysis.
The are several ways to estimate the offset vector:

\begin{itemize}
  \item Use gene counts (gene mode):
  \begin{enumerate}
     \item Use only raw counts.
     \item Fit a generalized linear model from gen counts.	
  \end{enumerate}
  \item Use the sum of exonic bins from each gene (bin mode). Is important to
  note that in this mode some reads are counted more than one time, because the
  overlap more than one bin. This mode is not available for junctions.
\end{itemize}

EdgeR package also provides a separate framework to detect splicing events, this
rely on their \texttt{diffSpliceDGE} method. This method uses an strategy
similar to the offset methodology. ASpli allow its use with the method
\texttt{DUreportBinSplice}.

\subsubsection{Usage of \texttt{DUreport}, \texttt{DUreportBinSplice} and
\texttt{junctionDUreport} methods }

\texttt{DUreport} method performs DE estimation for genes and DU estimation for
bins in the same run. The call of this method is done by:

<<write, echo=TRUE, eval=FALSE>>=
du <- DUreport( counts,
                targets,
                minGenReads = 10, 
                minBinReads = 5,
                minRds = 0.05, 
                offset = FALSE, 
                offsetAggregateMode =  c( "geneMode", "binMode" )[2],
                offsetUseFitGeneX = TRUE,
                contrast = NULL, 
                forceGLM = FALSE, 
                ignoreExternal = TRUE, 
                ignoreIo = FALSE, 
                ignoreI = FALSE,
                filterWithConstrasted = FALSE,
                verbose = FALSE )
@

Where the accepted parameters are:
\begin{description}
  \item[counts] an ASpliCounts object.
  \item[targets] a targets data frame built as explained in 
  \secref{sec:targetsDef}.
  \item[minGenReads] minimum value for reads in any condition required for a
  gene to be considered as expressed (see \secref{sec:de} ). The default value
  is 10.
  \item[minBinReads] minimum value for reads in any condition required for a
  bin to be included in the analysis (see \secref{sec:du} ). The default value
  is 5.
  \item[minRds] minimum value for read density for genes and bins required to be
  considered in the analysis. The default value is 0.05.
  \item[offset] a logical value that indicates, if \texttt{TRUE}, that an offset
  vector should be estimated and used in the analysis. If \texttt{FALSE}, bin
  counts by gene counts adjustment is used instead. The default value is 
  \texttt{FALSE} ( see \secref{sec:du} )
  \item[offsetAggregateMode] a character value that specifies if gene mode or
  bin mode of offset estimation should be used. It is only relevant when
  \item[offset] argument is \texttt{TRUE} (see \secref{sec:du}). The default
  value is \texttt{geneMode}.
  \item[offsetUseFitGeneX] a logical value that indicates if
  the offset vector should be computed using a GLM. This is only relevant when \texttt{offset}
  argument is \texttt{TRUE} and \texttt{offsetAggregateMode} argument is
  \texttt{geneMode} (see \secref{sec:du}). The default value is \texttt{TRUE}.
  \item[contrast] is a vector representing the coefficients for each unique
  condition in the analysis. They are assigned in the order given by
  \texttt{getConditions} method (see \secref{sec:targetsDef}). The default value
  is \texttt{NULL} which correspond to a vector with value -1 in the first
  position, 1 in the second, and zero for the remaining positions. This is for
  the paired comparison of the second condition versus the first condition.
  \item[forceGLM] a logical value that indicates that fitting a GLM should be
  used in the analysis for paired comparison, instead of an extact test (used
  by default for this type of comparison). The default value if \texttt{FALSE}.
  \item[ignoreExternal] a logical value indicating that external bins should
  be ignored in the analysis. The default value is \texttt{TRUE}.
  \item[ignoreIo] a logical value indicating that original intron bins should
  be ignored in the analysis. The default value is \texttt{TRUE}.
  \item[ignoreI] a logical value indicating that only exons should be included
  in the analysis. The default value is \texttt{FALSE}.
  \item[filterWithConstrasted] a logical value that specifies that gene and
  bin filtering when more than two conditions are present, should be done only
  with the data of those conditions and ignore the others. The default value
  id \texttt{FALSE}, it is strongly recommended to do not change this value.
  \item[verbose] a logical value that indicates that detailed information
  about each step in the analysis will be presented to the user.
  
\end{description}

\texttt{DUreportBinSplice} method performs DE estimation for genes and DU
estimation for bins in the same run. The call of this method is done by:

<<export, echo=TRUE, eval=FALSE >>=
du <- DUreportBinSplice( counts, 
                         targets, 
                         minGenReads = 10, 
                         minBinReads = 5,
                         minRds = 0.05, 
                         contrast = NULL, 
                         forceGLM = FALSE, 
                         ignoreExternal = TRUE, 
                         ignoreIo = TRUE, 
                         ignoreI = FALSE,
                         filterWithContrasted = FALSE )
@

Where the accepted parameters are:
\begin{description}
  \item[counts] an ASpliCounts object.
  \item[targets] a targets data frame built as explained in 
  \secref{sec:targetsDef}.
  \item[minGenReads] minimum value for reads in any condition required for a
  gene to be considered as expressed (see \secref{sec:de} ). The default value
  is 10.
  \item[minBinReads] minimum value for reads in any condition required for a
  bin to be included in the analysis (see \secref{sec:du} ). The default value
  is 5.
  \item[minRds] minimum value for read density for genes and bins required to be
  considered in the analysis. The default value is 0.05.
  \item[contrast] is a vector representing the coefficients for each unique
  condition in the analysis. They are assigned in the order given by
  \texttt{getConditions} method (see \secref{sec:targetsDef}). The default value
  is \texttt{NULL} which correspond to a vector with value -1 in the first
  position, 1 in the second, and zero for the remaining positions. This is for
  the paired comparison of the second condition versus the first condition.
  \item[forceGLM] a logical value that indicates that fitting a GLM should be
  used in the analysis for paired comparison, instead of an extact test (used
  by default for this type of comparison). The default value if \texttt{FALSE}.
  \item[ignoreExternal] a logical value indicating that external bins should
  be ignored in the analysis. The default value is \texttt{TRUE}.
  \item[ignoreIo] a logical value indicating that original intron bins should
  be ignored in the analysis. The default value is \texttt{TRUE}.
  \item[ignoreI] a logical value indicating that only exons should be included
  in the analysis. The default value is \texttt{FALSE}.
  \item[filterWithContrasted] a logical value that specifies that gene and
  bin filtering when more than two conditions are present, should be done only
  with the data of those conditions and ignore the others. The default value
  id \texttt{FALSE}, it is strongly recommended to do not change this value.
  
\end{description}

\texttt{junctionDUreport} method performs DU estimation for junctions. This is
independent of the genes DE and bins DU analysis. However, the results can be
merged with existing genes and bins results. The call of this method is done by:

<<export, echo=TRUE, eval=FALSE >>=
du <- junctionDUReport(  counts, 
                         targets, 
                         appendTo = NULL, 
                         minGenReads = 10,
                         minRds = 0.05,
                         threshold = 5,
                         offset   = FALSE,
                         offsetUseFitGeneX = TRUE,
                         contrast = NULL,
                         forceGLM = FALSE )
@

Where the accepted parameters are:
\begin{description}
  \item[counts] an ASpliCounts object.
  \item[targets] a targets data frame built as explained in 
  \secref{sec:targetsDef}.
  \item[appendTo] an ASpliDU object to which the results will be attached.
  \item[minGenReads] minimum value for reads in any condition required for a
  gene to be considered as expressed (see \secref{sec:de} ). The default value
  is 10.
  \item[minRds] minimum value for read density for genes required to be
  considered as expressed. The default value is 0.05.
  \item[threshold] minimum number of junction supporting reads required to be
  present in at least one sample of every conditions in order to include the
  junction in the analysis.
  \item[offset] a logical value that indicates, if \texttt{TRUE}, that an offset
  vector should be estimated and used in the analysis. If \texttt{FALSE},
  junction counts by gene counts adjustment is used instead. The default value
  is \texttt{FALSE} ( see \secref{sec:du} )
  \item[offsetUseFitGeneX] a logical value that indicates if
  the offset vector should be computed using a GLM. This is only relevant when \texttt{offset}
  argument is \texttt{TRUE} (see \secref{sec:du}). The default value is \texttt{TRUE}.
  \item[contrast] is a vector representing the coefficients for each unique
  condition in the analysis. They are assigned in the order given by
  \texttt{getConditions} method (see \secref{sec:targetsDef}). The default value
  is \texttt{NULL} which correspond to a vector with value -1 in the first
  position, 1 in the second, and zero for the remaining positions. This is for
  the paired comparison of the second condition versus the first condition.
  \item[forceGLM] a logical value that indicates that fitting a GLM should be
  used in the analysis for paired comparison, instead of an extact test (used
  by default for this type of comparison). The default value if \texttt{FALSE}.
  
\end{description}

The three methods described above return an \texttt{ASpliDU} object. 
This object contains the results for genes DE, bins DU and junction DU, which
can be consulted through their respective accessor methods: \texttt{genesDE},
\texttt{binsDU} and \texttt{junctionsDU}. See next section
(\secref{sec:asplidu}) for details about the specific contents of the data
obtained with these methods.

\subsubsection{Contents of ASpliDU object}
\label{sec:asplidu}

ASpliDU accessor methods \texttt{genesDE}, \texttt{binsDU} and
\texttt{junctionsDU} returns a \texttt{data.frame} object with the corresponding
results.

\begin{itemize}
  \item Results for genes differential expression analysis can be accessed with
  \texttt{genesDE}. This method returns a \texttt{data.frame}
  object which columns are specified below.
    \begin{description}
      \item[row.names] Gene name as is found in annotation.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section
      \secref{sec:binDefinition}).
      \item[locus\_overlap] Show the names of all other overlapping
      \textit{loci}.
      \item[gene\_coordinates] Show the genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[start] Left-most position of the gene.
      \item[end] Right-most position of the gene.
      \item[length] Number of bases covering the gene.
      \item[effective\_length] Number of bases that overlaps with annotated exons.
      \item[logFC] Show the observed change of expression level in a $log_{2}$
      scale.
      \item[pvalue] The pvalue for the DE test.
      \item[gen.fdr] The pvalue adjusted by \textbf{Benjamini-Hochberg} method.
      \\
    \end{description}
  \item Results for experimental junctions assuming that they correspond to
  introns involved in intron retention events can be accessed with
  \texttt{junctionPIR} method. The returning value is a \texttt{data.frame} with
  columns:
    \begin{description}
      \item[row.names] Bin name as given in section \ref{sectionBinG}:\nameref{sectionBinG}.
      \item[feature] Is the kind of bin: \textbf{E} for exonic bins, \textbf{I}
      for intronic bins and \textbf{Io} for introns before splitting.
      \item[event] Splicing event asigned to the bin (see section
      \ref{sec:binDefinition}:\nameref{sec:eventAssign})
      \item[locus] The name of the locus that contains the bin.
      \item[locus\_overlap] Show the names of all other overlapping \textit{loci}.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section \ref{sec:binDefinition}:\nameref{sec:binDefinition}).
      \item[gene\_coordinates] Show the genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[start] Left-most position of the bin.
      \item[end] Right-most position of the bin.
      \item[length] Number of bases covering the bin.
      \item[logFC] Show the observed change of usage in a $log_{2}$
      scale.
      \item[pvalue] The pvalue for the DU test.
      \item[gen.fdr] The pvalue adjusted by \textbf{Benjamini-Hochberg} method.
      \\
    \end{description}
  \item Results for experimental junctions assuming that the bearing
  exons are involved in alternative splice events can be accessed with
  \texttt{junctionPSI} method. The returning value is a \texttt{data.frame} with
  columns:
    \begin{description}
      \item[row.names] Junction name in format \texttt{chromosome.5'site-3'site}.
      \item[junction] If junction coincides with a junction inferred
      from the annotation, the name is shown as is given in section
      \ref{sectionBinG}:\nameref{sectionBinG}, otherwise contains \texttt{noHit}.
      \item[gene] The name of the locus that contains the junction.
      \item[strand] Sense strand of the gene.
      \item[multipleHit] Contains \texttt{yes} if junction spans multiple genes.
      \item[symbol] An optional common name for the gene, that must be provided at
      the moment of feature extraction (see section \ref{sec:binDefinition}:\nameref{sec:binDefinition}).
      \item[gene\_coordinates] Show the genomic coordinates of the gene with
      format \texttt{chromosome:start-end}.
      \item[bin\_spanned] The names of the bins spanned by the junction.
      \item[j\_within\_bin] If junction fall within a single bin, the name of
      that bin is shown.
      \item[junction count data] One column per sample containing raw junction
      counts if \texttt{offset} argument is \texttt{TRUE}, or junction counts
      adjusted by gene counts otherwise.
      \item[logFC] Show the observed change of usage in a $log_{2}$
      scale.
      \item[pvalue] The pvalue for the DU test.
      \item[gen.fdr] The pvalue adjusted by \textbf{Benjamini-Hochberg} method.
      \item[bin\_start\_hit] List of junction names that shares the start
      position with current junction, in format \texttt{chromosome:start-end}.
      \item[junction sharing start data] One column per samples containing
      the counts of junctions sharing the start with the current junction.
      Junction counts can be raw or adjusted following the same criteria
      described for \texttt{junction count data} column.
      \item[junction sharing start ratio] One column by condition containing the
      ratio of current junction among all junctions sharing their start
      position (see \secref{sec:psir}). The ratio is calculated using raw
      or adjusted counts. 
      \item[end\_start\_hit] List of junction names that shares the start
      position with current junction, in format \texttt{chromosome:start-end}.
      \item[junction sharing end data] One column per samples containing
      the counts of junctions sharing the end with the current junction.
      Junction counts can be raw or adjusted following the same criteria
      described for \texttt{junction count data} column.
      \item[junction sharing end ratio] One column by condition containing the
      ratio of current junction among all junctions sharing their end
      position (see \secref{sec:psir}). The ratio is calculated using raw
      or adjusted counts. 
      \\
    \end{description}  

\end{itemize}

\subsection{Output and results}
\label{sec:output}

At each module, results are stored in \texttt{ASpliObjects}. Self-explanatory
tables can be exported at each step of the analysis. Using \texttt{write} 
functions, it is possible to export tab delimited tables in a features-level 
output folder:

<<write, echo=TRUE, eval=FALSE>>=
writeCounts(counts, "example_counts")
writeDU(du, output.dir="example_du");
writeAS(as=as, output.dir="example_as");
writeAll(counts=counts, du=du, as=as, output.dir="example_all")
@

Column of text tables has the same name and meaning that those described in
\secref{sec:countsContents}, \secref{secASpliAsContent} and
\secref{sec:asplidu}. Exported text tables files are arranged in subfolders 
"exons", "genes", "introns" and "junctions". \texttt{writeAS} creates an 
additional text table:
\begin{description}
  \item[as\_discovery.tab] Contains junction counts and PSI/PIR metrics for all
    bins.
\end{description}

\texttt{writeAll} methods writes some additional tables:
\begin{description}
  \item[summary.tab] Contains bin DU usage for all bins together.
  \item[bins\_du\_psi\_pir.tab] Contains bin DU usage for all bins merged
  with PSI/PIR metrics.
\end{description}

\subsection{Plots}
\label{plots}

Inspection of coverage plots is a good practice for the interpretation of 
analytical results. After selection of AS candidates it is possible to plot the 
results in a genome browser manner highlighting  alternatively spliced regions 
using the function \texttt{plotGenomicRegions}. Coverage data is taken from
\textbf{bam} files. \texttt{plotGenomicRegions} function will draw a coverage
plot for each condition, not for each sample, therefore coverage data from
samples needs to be merged. \texttt{plotGenomicRegions} can extract and merge
the reads of the corresponding \textit{locus} of interest on-the-fly from
sample's \textbf{bam} files or can use preiviously merged \textbf{bam} generated
by the user. When the number of \textit{loci} to be plotted is small (ten or
less) the first approach is usually faster, otherwise is better to merge and
index \textbf{bam} files beforehand. \textbf{Bam} files can be merged using
\texttt{samtools} utility or \texttt{RSamtools} package. 

<<targetsPlot, echo=TRUE, eval=FALSE>>=
library(GenomicFeatures)
bamfiles <- system.file( 'extdata', 
    c('A_C_0.bam', 'A_C_1.bam', 'A_C_2.bam',
      'A_D_0.bam', 'A_D_1.bam', 'A_D_2.bam',
      'B_C_0.bam', 'B_C_1.bam', 'B_C_2.bam',
      'B_D_0.bam', 'B_D_1.bam', 'B_D_2.bam' ),
    package = "ASpli") 

targets <- data.frame(
  row.names = c( 'A_C_0', 'A_C_1', 'A_C_2',
                 'A_D_0', 'A_D_1', 'A_D_2',
                 'B_C_0', 'B_C_1', 'B_C_2',
                 'B_D_0', 'B_D_1', 'B_D_2' ),
  bam = bamfiles,
  f1  = rep( c('A','B'), each=6 ),
  f2  = rep( c('C','D'), 2, each=3 ),
  stringsAsFactors = FALSE )

genomeTxDb <- makeTxDbFromGFF( system.file( 'extdata', 'genes.mini.gtf',
package='ASpli'))
features <- binGenome( genomeTxDb )

# Draw a single plot on a window
# xIsBin arguments is used to specify if names given are bins or genes.
# When multiple names are given, all must correspond to bins or correspond to
# genes. Mixed names are not allowed.

plotGenomicRegions( features, 'GENE02:E002' , genomeTxDb, targets, 
    xIsBin = TRUE, verbose = TRUE)

# Draw a single plot on a window, with a custom layout
plotGenomicRegions( features, 'GENE02:E002' , genomeTxDb, targets, xIsBin =
   TRUE, layout = matrix( c('A_C','A_D','B_C','B_D') ,nrow = 1 ), 
   verbose = TRUE, color = '#AA8866')

# Draw many plots on files.
# Argument outfileType define the file format of the generated files.
# Argument useTransparency specified that transparency can be used to draw the
# plot. No all graphic devices support transparency.
# Argument annotationHeight specifies the proportional height of the plot used
# to draw the annotation.
# Argument deviceOpt gives additional arguments (width and height in this
# example) to the underlying graphic device (pdf in this example)
# One of the plots generated here is shown in a figure below.
plotGenomicRegions( features, c( 'GENE02:E002', 'GENE01:E002', 'GENE02:E003' ), 
    genomeTxDb, targets, xIsBin = TRUE, outfolder = 'grplots', verbose = TRUE,
    color = '#334466', outfileType='pdf', useTransparency = TRUE,
    annotationHeight = 0.12, deviceOpt = c( width = 8, height = 6) )
    
# Draw a single plot on a window, with premerged bams
mergedBams <- data.frame(
  row.names = c( 'A_C','A_D','B_C','B_D' ),
  bam = c( 'A_C.bam',  # Warning, these files do not exists.
           'A_D.bam',  # Bam files should be indexed.
           'B_C.bam',
           'B_D.bam' ))
plotGenomicRegions( features, 'GENE02:E002' , genomeTxDb, targets,
   xIsBin = TRUE, verbose = TRUE, preMergedBAMs = mergedBams )    
@

Figure \ref{fig:plotGenomicRegion} shows a plot from one of the previuos
examples. 

\begin{figure}[ht!]
  \begin{center}
  \includegraphics[width=0.55\textwidth]{images/GENE02_E002_gr.pdf}
  \end{center}
  \caption{Genomic plot of a gene containing a exon skipping splicing event
  highlighted for four different conditions.}
  \label{fig:plotGenomicRegion}
\end{figure}

In addition to coverage plots, \texttt{ASpli} is able to draw plots containing
raw gene counts, raw bin counts, PSI/PIR value, inclusion and exclusion junction
counts for a single bin in all, or a subset, conditions of the analysis. This
plot is useful to integrate evidences of different sources in order to explain
an experimental result. The function to make these plots is \texttt{plotBins},
below there are some examples of use.

Example 1:
<<plotbins, echo=TRUE, eval=FALSE>>=
# Defines an experiment with one experimental factor (genotype, with two
# values: wild-type (WT) and mutant (MT) ), and three replicate samples for each
# condition.
targets <- data.frame(
  row.names = c( 'WT1', 'WT2', 'WT3', 
                 'MT1', 'MT2', 'MT3' ),
  bam = c( 'WT1.bam', 'WT2.bam', 'WT3.bam', 
           'MT1.bam', 'MT2.bam', 'MT3.bam' ),
  genotype = c( 'WT','WT','WT','MT','MT','MT' ),
  stringsAsFactors = FALSE )
# Specifies what factors and values will be plotted, in this example all of
# them.
fv = list( genotype = c( 'WT', 'MT' ) ) 
plotbins(
        counts, 
        as,
        'GENE02:E002', 
        factorsAndValues = fv, 
        targets )
@

Example 2:
<<plotbins, echo=TRUE, eval=FALSE>>=
# Defines an experiment with two experimental factor (treat, with four
# values: A, B, C and D; and time, with twelve values from T1 to T12.), and
# two replicate samples for each condition. In the definition, many values are
# not shown to reduce the space of the example, and are replaced by '...' . 
targets <- data.frame(
  row.names = c( 'A_T1.r1', 'A_T1.r2', 
                 'A_T2.r1', 'A_T2.r2',
                 ... , 
                 'D_T12.r1', 'D_T12.r2' ),
  bam = c( 'A_T1.r1.bam', 'A_T1.r2.bam', 
           'A_T2.r1.bam', 'A_T2.r2.bam',
           'D_T12.r1.bam', 'D_T12.r2.bam' ),
  treat = c( 'A', 'A',
             'A', 'A',
             ... ,
             'D', 'D' ),
  time = c( 'T1', 'T1',
            'T2', 'T2',
            ... ,
            'T12', 'T12' ),
  stringsAsFactors = FALSE )
# Draw the plots.
plotbins(
        counts, 
        as,
        'GENE02:E002', 
        factorsAndValues = fv, 
        targets )

# Specifies what factors and values will be plotted. In this example there are
# two factor, the first one in 'fv' list is the main factor, values will
# be grouped by this factor. 
fv = list( time = c( 'T1', 'T2', ... , 'T12' ) ,
           treat = c( 'A', 'B', 'C', 'D' ) ) 
# Draw the plots
plotbins(
        counts, 
        as,
        'GENE02:E002', 
        factorsAndValues = fv, 
        targets )
@

Figures \ref{fig:plotbinEx1} and \ref{fig:plotbinEx2} shows the images that can
be obtained in setups like shown in the examples above. Note that one image 
contains barplots, and the other contains lines. They are selected automatically
according to the number of conditions and values per condition. However, it is 
possible to force to one of them setting the argument \texttt{useBarPlots} in
\texttt{plotBins} method.

\begin{figure}[ht!]
  \centering
  \includegraphics[width=0.6\textwidth]{images/GENE02_E002_pr.pdf}
  \caption{Plots of bin gene, and junction counts, and PSI/PIR metric for a
  single bin of an experiment containing a time course of 12 posints in four
  conditions. Gene expression seem to be constant across all conditions,
  however, the usage of the bin oscilates with time. Inclusion and
  exclusion junctions show a concordant behaviour.}
  \label{fig:plotbinEx1} 
\end{figure}

\begin{figure}[ht!]
  \centering
  \includegraphics[width=0.25\textwidth]{images/GENE02_E002b_pr.pdf}
  \caption{Plots of bin gene, and junction counts, and PSI/PIR metric for a
  single bin of an experiment containing two combinatorial factors. Of those,
  only two conditions are shown. }
  \label{fig:plotbinEx2} 
\end{figure}

\subsection{Utility functions}

ASpli includes many functions that help the use to manipulate and interrogate
the different ASpli objects.

\subsubsection{Consulting data in \texttt{ASpliDU} objects}

ASpliDU objects can contain data for gene, bins and/or junctions depending on
what function was used to create it. \texttt{DUreport} and
\texttt{DUreportBinSplice} methods returns \texttt{ASpliDU} objects with data
for gene and bins. \texttt{junctionDUreport} function returns an object with
junction data only or with gene, bins and junction depending on the input.
There are two function that are useful to know what kind of data an
\texttt{ASpliDU} contains: \texttt{containsJunctions} and
\texttt{containsGenesAndBins}. 

<<containsDUfunctions, echo=TRUE, eval=FALSE>>=
du <- aspliDUexample1()
containsJunctions( du )
containsGenesAndBins( du )
@

\subsubsection{Filtering \texttt{ASpliCount} objects}

Genes, bins and junctions present in an \texttt{ASpliCount} object can be
filtered by abundance according to several statments:
\begin{itemize}
  \item number of counts or read density.
  \item mean value or minimum value.
  \item grouping by condition or on the whole set.
  \item all conditions must meet the selection criteria or at least one
  condition must do it.
\end{itemize}

\texttt{filterReadCounts} method allow this filtering. Here there are some
examples:

%<<filterReadCounts, echo=TRUE, eval=FALSE>>=
%filterReadCounts(du)
%@

\subsubsection{Filtering \texttt{ASpliDU} objects}
Results present in \texttt{ASpliDU} can be filtered by \textbf{logFC} and
\textbf{FDR} with \texttt{filterDU} function. Elements with and \textbf{FDR}
value lower or equal to a given threshold, and with an absolute \textbf{logFC}
greater or equal to a given value are kept. Default values for \textbf{logFC} and
\textbf{FDR} threshold values are 0 and 1 respectively, in other words, is not
filtered by that value.

<<filterDU1, echo=TRUE, eval=FALSE>>=
# Filter genes that changes at least a 50 percent their expression with a 
# FDR value of 0.05 or less, and bins changes at least a 50 percent their expression with a 
# FDR value of 0.05 
du <- aspliDUexample1()
duFilt <- filterDU(du, 'genes', '0.05', log(1.5,2))
duFilt <- filterDU(duFilt, 'bins', '0.1', log(1.5,2))
@

It is possible to filter only those elements that has a positive
(or negative) \textbf{logFC} using the arguments \texttt{absLogFC} and
\texttt{logFCgreater}. \texttt{absLogFC} forces to take the absolute value of
\textbf{logFC}, this is \texttt{TRUE} by default. \texttt{logFCgreater}
specifies, if \texttt{TRUE}, that the \textbf{logFC} value must be greater than
the threshold.

<<filterDU2, echo=TRUE, eval=FALSE>>=
# Filter genes that reduces their expression to the half
du <- aspliDUexample1()
duFilt <- filterDU( du, 'genes', '0.05', -1, absLogFC = FALSE,
                    logFCgreater = FALSE  ) 
@

\subsubsection{Subsetting \texttt{ASpli} objects}

\texttt{ASpliCounts}, \texttt{ASpliAS}, targets \texttt{data.frame} and list of
\texttt{GAlignments} (i.e. bam files data) can be subsetted by conditions or
samples.
<<subset, echo=TRUE, eval=FALSE >>=
# Subset counts
counts <- aspliCountsExample()
countsSmall <- subset(counts, targets, select = c("A_C", "A_D") )
# Subset AS results. Note that PIR/PSI metrics are not recalculated.
as <- aspliASexample()
asSmall <- subset(as, targets, select = c("A_C_0", "A_C_1") )
# Subset targets
targets      <- aspliTargetsExample()
targetsSmall <- subsetTargets( targets, c("A_C", "A_C") )
# Subset BAMs
bams     <- aspliBamsExample()
bamSmall <- subsetBams( bams, targets, c("A_C", "A_C") )
@

\subsubsection{Merging DU and junction data for bins}

\texttt{mergeBinDUAS} function merge bin data from DU analysis contained in an
\texttt{ASpliDU} object and bin data from junction repertoire usage and PSI/PIR
metrics contained in an \texttt{ASpliAS} object. Also a delta PSI/PIR value is
calculated from a contrast vector.

<<mergeBinDUAS, echo=TRUE, eval=FALSE>>=
du <- aspliDUexample1()
as <- aspliASexample()
mergeBinDUAS(du,as, targets)
@

\section{Example data}

\subsubsection{Synthetic data set}
ASpli contains a small synthetic data set to test de complete working pipeline.
It consist of a genome with ten genes with multiple isoforms, a set of
\textbf{BAM} files from an experiment with two$\times$two factorial design, and
a \textbf{GTF} file with the corresponding annotation. The two factors in this
examples are called \textbf{f1} and \textbf{f2}. \textbf{f1} can have value
\textbf{A} or \textbf{B} and \textbf{f2} can have value \textbf{C} or
\textbf{D}, defining four conditions: \textbf{A.C}, \textbf{A.D}, \textbf{B.C}
and \textbf{B.D}. There are three replicates for each conditition (see table
\ref{tab:Ex01} ) .

\begin{table}[H]
  \begin{center}
    \begin{tabular}{lllll}
    Sample   & f1 & f2 & replicate & condition \\ \hline
    1        & A  & C  & 0         & A.C       \\
    2        & A  & C  & 1         & A.C       \\
    3        & A  & C  & 2         & A.C       \\
    4        & A  & D  & 0         & A.D       \\
    5        & A  & D  & 1         & A.D       \\
    6        & A  & D  & 2         & A.D       \\
    7        & B  & C  & 0         & B.C       \\
    8        & B  & C  & 1         & B.C       \\
    9        & B  & C  & 2         & B.C       \\
    10       & B  & D  & 0         & B.D       \\
    11       & B  & D  & 1         & B.D       \\
    12       & B  & D  & 2         & B.D       \\
    \end{tabular}
  \end{center}
  \caption{Experimental design of the example data set.}
  \label{tab:Ex01}
\end{table}

The firts step of the workflow is to load the annotation from the reference
sequence and extract their features.

<<Ex01.a, echo=TRUE, eval=FALSE>>=

library( GenomicFeatures )
# gtfFileName contains the full path to the example gtf file in your system.
gtfFileName <- aspliExampleGTF()

# Create a TxDb object using makeTxDbFromGFF function from GenomicFeatures 
# package
genomeTxDb <- makeTxDbFromGFF( gtfFileName )

# Extract the genomic features
features <- binGenome( genomeTxDb )

@

Also, the exprimental sequencing data must be loaded, and the experimental
samples, and factors should be defined.

<<Ex01.b, echo=TRUE, eval=FALSE>>=

# bamFiles contains the full path of the bam files in your system
bamFiles <- aspliExampleBamList( )

# Define the targets
targets <- data.frame( 
  row.names = paste0('Sample',c(1:12)),
  bam = bamFiles,
  f1 = c( 'A','A','A','A','A','A',
          'B','B','B','B','B','B'),
  f2 = c( 'C','C','C','D','D','D',
          'C','C','C','D','D','D'),
  stringsAsFactors = FALSE
)

# Examinate the condition names
getConditions(targets)

# Load the bam files
bams = loadBAM(targets)
@

After the targets are defined and the sequence data is read, you can continue to
assign the number of counts to each feature.

<<Ex01.c, echo=TRUE, eval=FALSE>>=
counts <- readCounts( features, bams, targets, readLength = 100, 
                      maxISize = 5000 )
@

Now you can run the DE/DU analysis, here will be avaualted a test to hypothesis
that the expression of some genes respond to the interaction of \textbf{f1} and
\textbf{f2}. In other words, that the observed effect of applying the two
factors can not be explained by the independent contribution of each factor.
The interaction can be expressed as:
\begin{equation}
I = ( B.D - B.C ) - ( A.D - A.C ) 
\end{equation}
The coefficients of the terms in this expression can be represented in the
contrast vector $[1,-1,-1,1]$, assuming that they correspond to the order given
by \texttt{getConditions} function.

<<Ex01.d, echo=TRUE, eval=FALSE>>=
# DU/DE analysis for genes and bins.
du <- DUreport( counts, targets, contrast = c( 1, -1, -1, 1 ) )
# DU analysis for junctions
du <- junctionDUReport( counts, targets, appendTo = du, 
                        contrast = c( 1, -1, -1, 1 ) )
@

The discovery of AS events can be done with:
<<Ex01.e, echo=TRUE, eval=FALSE>>=
# AS analysis 
as <- AsDiscover( counts, targets, features, bams, readLength = 100)
@

Finally, you can select your top tags integrating differential usage data and
junction data.

<<Ex01.f, echo=TRUE, eval=FALSE>>=
# Select top tags from DU
topTagsBins <- filterDU( du, what = 'bins', fdr = 0.05, logFC = log(1.5,2) )
# DU results are merged with AS results
topTagsBins <- mergeBinDUAS( du, as, targets, contrast = c( 1, -1, -1, 1 ) )
# We filter those top tags that have also junction evidence that supports the
# differential usage, a change of at least 0.1 between observed and expected 
# in the PSI or PIR metric is required.
topTagsBins <- topTagsBins[ abs(topTagsBins[,'delta' ]) > 0.10,]
@

Bin \textbf{GENE10:E002} is the one with the lowest \textbf{FDR} value, now we
can examine it in a plot. The results in figure \ref{fig:Ex01gr} show that a
change in \textbf{f1} from \textbf{A} to \textbf{B} has not the same outcome for
\textbf{C} or \textbf{D} in \textbf{f2}. 

<<Ex01.f, echo=TRUE, eval=FALSE>>=
plotGenomicRegions( features,'GENE10:E002', genomeTxDb, targets )
@

\begin{figure}[!ht]
  \centering
  \includegraphics[width=0.55\textwidth]{images/GENE10_E002_gr.pdf}
  \caption{Genomic region of example bin \textbf{GENE10:E002}.}
  \label{fig:Ex01gr}
\end{figure}

\subsubsection{Real data set}

It is possible to run a demo of \texttt{ASpli} using public \RNAseq data 
available via Bioconductor. It requieres installing of 
\texttt{RNAseqData.HNRNPC.bam.chr14} package and loading  an small TranscriptDb
of Human Genome only for chromosome 14. This subset of reads is part of an
experiment intended to analyze splicing factors and their relationship with 
exonization of Alu elements \cite{Zarnack2013453}. We load the package and the 
genome:
%<<librariesEx, echo=TRUE, eval=TRUE>>=
<<Ex02.a, echo=TRUE, eval=FALSE>>=
library(RNAseqData.HNRNPC.bam.chr14)
@

In this case we are loading a genome \texttt{TranscriptDb} already created and 
available in the \texttt{ASpli} package for the \textit{Homo sapiens}
chromosome 14:

%<<loadDb, echo=TRUE, eval=TRUE>>=
<<Ex02.b, echo=TRUE, eval=FALSE>>=
chr14 <- system.file("extdata","chr14.sqlite", package="ASpli")
genome <- loadDb(chr14)
@

And now we are ready to start the complete analysis.
%<<binGenome, echo=TRUE, eval=TRUE>>=
<<Ex02.c, echo=TRUE, eval=FALSE>>=
features <- binGenome(genome) 
@

This data set contains data for of two conditions \textbf{KD} and \textbf{CT}
with four replicates each one. 
Create targets object using the information of the available files:
%<<targetsEx, echo=TRUE, eval=TRUE>>=
<<Ex02.d, echo=TRUE, eval=FALSE>>=
targets <- data.frame( 
              bam = RNAseqData.HNRNPC.bam.chr14_BAMFILES,
              treat = c( "CT", "CT", "CT", "CT", 
                         "KD", "KD", "KD", "KD" ),
              stringsAsFactors = FALSE )
@
Then, load the bam files:
%<<loadBam, echo=TRUE, eval=TRUE>>=
<<Ex02.e, echo=TRUE, eval=FALSE>>=
bam <- loadBAM(targets)
@
Overlap alignments against features:
%<<todosJuntos, eval=TRUE>>=
<<Ex02.f, eval=FALSE>>=
counts <- readCounts( features, bam, targets, readLength=100L, maxISize=50000 )
@
Perform DE/DU analysis
%<<DU, eval=TRUE>>=
<<Ex02.g, eval=FALSE>>=
du <- DUreport(counts, targets)
@
Analyze AS using junctions:
%<<AS, eval=TRUE>>=
<<Ex02.g, eval=FALSE>>=
as <- AsDiscover (counts, targets, features, bam, readLength=100L, threshold = 5, cores = 1 )
@

Select top tags, using DU and AS information:
%<<topTags, echo=TRUE, eval=TRUE>>=
<<Ex02.h, echo=TRUE, eval=FALSE>>=
duFiltered <- filterDU( du, 'bins', fdr = 0.05, logFC = log(1.5,2) ) 
merged     <- mergeBinDUAS( duFiltered, as, targets )
@

Bin ZNF410:E013, from top tags, has a usage change that is statistically
significant and the change of PSI/PIR metric is greater than 0.3. We can inspect
it visually, see figure \ref{Ex02.plot}.

%<<targetsPlot, eval=TRUE>>=
<<Ex02.i, eval=FALSE>>=
# Show the complete gene, with the bin hightlighted
plotGenomicRegions( features, 'ZNF410:E013', genome, targets, verbose=TRUE,
                    colors = matrix(c('#223344','#223344'), ncol=2 ), outfileType='pdf', outfolder='grplots')
# Show a zoom in in the bin
plotGenomicRegions( features, 'ZNF410:E013', genome, targets, verbose=TRUE,
                    colors = matrix(c('#223344','#223344'), ncol=2 ),
                    zoomOnBins = 0.05)
@
\begin{figure}[ht!]
\centering
\includegraphics[width=0.85\textwidth]{images/ZNF410_E013_gr.pdf}
\caption{Plot of ZNF410:E013 exon skipping}
\label{Ex02.plot}
\end{figure}

\bibliography{ASpli}
\end{document}