%\VignetteIndexEntry{crlmmDownstream}
%\VignetteKeywords{copy number, genotype, SNP}
%\VignettePackage{VanillaICE}
\documentclass{article}
\usepackage{amsmath}
\usepackage{graphicx}
\usepackage[numbers]{natbib}
\usepackage{color}
\usepackage[margin=1in]{geometry}

\newcommand{\scscst}{\scriptscriptstyle}
\newcommand{\scst}{\scriptstyle}
\newcommand{\Rpackage}[1]{\textit{#1}}
\newcommand{\Rfunction}[1]{\texttt{#1}}
\newcommand{\Robject}[1]{\texttt{#1}}
\newcommand{\Rclass}[1]{\texttt{#1}}
\newcommand{\R}{\textsf{R}}
\newcommand{\hmmoptions}{\Robject{HmmOptions}}
\newcommand{\hmmparam}{\Robject{HmmParameter}}
\newcommand{\crlmm}{\Rpackage{crlmm}}
\newcommand{\oligo}{\Rpackage{oligo}}

\title{Integration with the crlmm package for copy number inference}
\author{Robert Schgarpf}

\begin{document}

\maketitle

We load a portion of chromosome 8 from 2 HapMap samples that were
processed using the \Rpackage{crlmm} package.

<<loadData>>=
library(oligoClasses)
library(VanillaICE)
library2(crlmm)
library2(SNPchip)
library2(IRanges)
data(cnSetExample, package="crlmm")
@

The data \Robject{cnSetExample} is an object of class \Rclass{CNSet}.
We coerce the \Rclass{CNSet} object to a list class that contains
information on copy number (log R ratios), genotypes, genotype
probabilities, and B allele frequencies.

<<coerce2OligoSnpSet>>=
oligoList <- constructOligoSetListFrom(cnSetExample)
@

The \verb+[[+ method can be used to extract a \Rclass{oligoSnpSet} for
the first element in the list:

<<subset>>=
oligoSet <- oligoList[[1]]
@

Next, we fit a 6-state hidden markov model from estimates of the B
allele frequency and log R ratios.

<<hmm>>=
res <- hmm(oligoSet, p.hom=0.1, nupdates=5, TAUP=1e8)
@

The \texttt{TAUP} parameter scales the transition probability matrix.
Larger values of \texttt{TAUP} makes it more expensive to transition
from the normal copy number state to states with altered copy
number. In the following code chunk, we use a lattice multi-panel
display to plot each of the altered stated.  We frame each alteration
by plotting a genomic interval of 200kb on each side (using the
\texttt{frame=200e3} argument):

<<xyplot>>=
rd <- res[chromosome(res) == "chr8", ]
rd <- res[!state(res)%in%c(3,4), ]
if(require(SNPchip)){
	fig <- xyplotLrrBaf(rd, oligoSet,
			    frame=200e3,
			    panel=xypanelBaf,
			    scales=list(x="free"),
			    par.strip.text=list(cex=0.9),
			    cex=0.4,
			    state.col="black",
			    state.cex=0.8,
			    pch=21)
}
@
%
%<<echo=FALSE>>=
%pdf("crlmmDownstream-latticeFig.pdf", width=10, height=7)
%@
%
<<latticeFig, fig=TRUE,include=FALSE, width=12, height=6>>=
print(fig)
@
%
%<<echo=FALSE>>=
%dev.off()
%@


\begin{figure}[t]
  \centering
  \includegraphics[width=\textwidth]{crlmmDownstream-latticeFig}
  \caption{\label{fig:chr8} Plot of log R ratios (grey) and B allele
    frequencies (blue). The B allele frequencies have a range of 0-1
    and were rescaled for ease of viewing alongside the log R
    ratios. Each panel displays one region with a copy number
    alteration predicted from the 6-state HMM with padding on either
    side given by the \texttt{frame} argument.}
\end{figure}


\section{Session Information}

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

<<echo=FALSE, results=tex>>=
toLatex(sessionInfo())
@

\end{document}