%\VignetteIndexEntry{BonEV}
\documentclass[11pt]{article}

\usepackage{amsmath,epsfig,fullpage,hyperref}
\usepackage{underscore}
\parindent 0in

%\newcommand{\Robject}[1]{{\texttt{#1}}}
%\newcommand{\Rfunction}[1]{{\texttt{#1}}}
%\newcommand{\Rpackage}[1]{{\textit{#1}}}

\begin{document}
\SweaveOpts{concordance=TRUE}

\title{\bf Introduction to BonEV package}

\author{Dongmei Li}

\maketitle

\begin{center}
Clinical and Translational Science Institute, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642-0708\\
\end{center}

\tableofcontents

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Overview}

This document provides an introduction to the {\tt BonEV} package. The {\tt BonEv} package calculates the adjusted P-values from user-provided raw P-values through the Bon-EV multiple testing procedure that controls the false discovery rates at user-defined level alpha. The Bon-EV multiple testing procedure is developed based on the Bonferroni procedure with integrated estimates from the Benjamini-Hochberg procedure and the Storey's q-value procedure. It controls false discovery rates through controlling the expected number of false discoveries.


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Getting started}

The {\tt BonEV} package can be installed and loaded through the following R code.

Install the {\tt BonEV} package with:
%code 1
<<eval=FALSE, echo=TRUE>>=
install.packages("BonEV")
@

Load the {\tt BonEV} package with:

%code 2
<<eval=FALSE, echo=TRUE>>=
library(BonEV)
@

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{{\tt Bon_EV} function}

There is one function in the {\tt BonEV} package: {\tt Bon_EV}. The function requires raw P-values in the vector format and user-defined alpha level for false discovery rates control. {\tt Bon_EV} will generate adjusted P-values from the Bon-EV multiple testing procedure that is developed based on the Bonferroni procedure with integrated estimates from the Benjamini-Hochberg procedure and the Storey's q-value procedure. {\tt Bon_EV} controls false discovery rates through controlling the expected number of false discoveries.

The following is an example using the {\tt Bon_EV} function. The raw P-values in the hedenfalk data set from the qvalue package are used as the input to get adjusted P-values from the Bon-EV multiple testing procedure with the false discovery rate controlled at level alpha = 0.05. Then, the adjusted P-values from the Bon-EV multiple testing procedure are compared with adjusted P-values obtained from the Benjamini-Hochberg and Storey's q-value procedures.

%code 3
<<eval=TRUE, echo=TRUE>>=
library(qvalue)
data(hedenfalk)
library(BonEV)
pvalues <- hedenfalk$p
adjp <- Bon_EV(pvalues, 0.05)
summary(adjp)
results <- cbind(adjp$raw_P_value, adjp$BH_adjp, adjp$Storey_adjp, adjp$Bon_EV_adjp)
colnames(results) <- c("raw_P_value", "BH_adjp", "Storey_adjp", "Bon_EV_adjp")
results[1:20,]
summary(results)

##Compare with Benjami-Hochberg and Storey's q-value procedures
sum(adjp$raw_P_value <= 0.05)
sum(adjp$BH_adjp <= 0.05)
sum(adjp$Storey_adjp <= 0.05)
sum(adjp$Bon_EV_adjp <= 0.05)
@

\end{document}