[Genabel-commits] r1909 - pkg/ProbABEL/doc
noreply at r-forge.r-project.org
noreply at r-forge.r-project.org
Sun Jan 4 23:45:14 CET 2015
Author: lckarssen
Date: 2015-01-04 23:45:14 +0100 (Sun, 04 Jan 2015)
New Revision: 1909
Modified:
pkg/ProbABEL/doc/ProbABEL_manual.tex
Log:
Improvements to the ProbABEL documentation:
- Uncomment section on the --interaction_only option
- Use \lstinline{} instead of \texttt{} when referring to command line options, otherwise the two -- are turned into an n-dash by LaTeX
- Increase the default font size to the standard size (instead of footnotesize) so that the \lstinline{} text is the same size as the regular text. As a result some listing environments have to scale down their fonts to footnotesize in order to be more readable.
Modified: pkg/ProbABEL/doc/ProbABEL_manual.tex
===================================================================
--- pkg/ProbABEL/doc/ProbABEL_manual.tex 2015-01-04 22:12:40 UTC (rev 1908)
+++ pkg/ProbABEL/doc/ProbABEL_manual.tex 2015-01-04 22:45:14 UTC (rev 1909)
@@ -24,6 +24,7 @@
\usepackage{verbatim}
\usepackage{listings}
+\lstMakeShortInline{|}
\lstloadlanguages{Bash}
\definecolor{lstbgcolor}{rgb}{0.9,0.9,0.9}
\lstset{
@@ -44,7 +45,7 @@
keywordstyle=\color[rgb]{0,0,1},
commentstyle=\color[rgb]{0,0.4,0},
stringstyle=\color[rgb]{0.5,0,1},
- basicstyle=\footnotesize\ttfamily,
+ basicstyle=\ttfamily,
backgroundcolor=\color{lstbgcolor},
}
@@ -160,7 +161,7 @@
\item If your computer runs Debian Linux\footnote{At the moment \PA{}
is only available in Debian testing and unstable.} (and you have
administrative rights on it), you can install ProbABEL like this:
- \begin{lstlisting}
+ \begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ apt-get install probabel
\end{lstlisting}
\item Zip files with pre-compiled binaries (if available) can be found
@@ -223,18 +224,18 @@
library, go to \url{http://eigen.tuxfamily.org} and download the
\texttt{tar.gz} file of the latest version of EIGEN (3.2.1 at the time
of writing). Extract the files:
-\begin{lstlisting}
+\begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ tar -xzf 3.2.1.tar.gz
\end{lstlisting}
This will create a directory called \texttt{eigen-eigen} followed by a
series of letters and digits. For simplicity we rename it to EIGEN
-\begin{lstlisting}
+\begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ mv eigen-eigen-6b38706d90a9 EIGEN
\end{lstlisting}
Now it's time to extract the \PA{} source code and move into the
directory that is created:
-\begin{lstlisting}
+\begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ tar -xzf probabel-0.5.0.tar.gz
user at server:~$ cd probabel-0.5.0
\end{lstlisting}
@@ -242,7 +243,7 @@
be found and where we want to install \PA{}. Let's install in a
subdirectory of your home directory,
e.g.~\texttt{/home/yourusername/ProbABEL}:
-\begin{lstlisting}
+\begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ ./configure \
--prefix=/home/yourusername/ProbABEL/ \
--with-eigen-include-path=/home/yourusername/EIGEN
@@ -259,9 +260,9 @@
to the \texttt{-j} option. For example for four cores run
\texttt{make -j4}.} The next step will check the compiled code,
after wich you install the program, documentation and examples to the
-directory you specified previously with the \texttt{--prefix} argument
+directory you specified previously with the \lstinline{--prefix} argument
to the \texttt{./configure} command.
-\begin{lstlisting}
+\begin{lstlisting}[basicstyle=\footnotesize\ttfamily,]
user at server:~$ make
user at server:~$ make check
user at server:~$ make install
@@ -353,7 +354,7 @@
(\texttt{.fvi} and \texttt{.fvd} files) in which case
\texttt{ProbABEL} will operate much faster, and in low-RAM mode
(approx.~128 MB). On the command line simply specify the \texttt{.fvi}
-file as argument for the \texttt{--dose} option
+file as argument for the \lstinline{--dose} option
(cf.~section~\ref{sec:runanalysis} for more information on the options
accepted by \texttt{ProbABEL}). See the R libraries \GA{} and
\DA{} on how to convert MaCH and IMPUTE files to
@@ -457,7 +458,7 @@
There are in total 11 command line options you can specify to the
\PA{} analysis functions \texttt{palinear} or \texttt{palogist}. If
-you run either program with the \texttt{--help} option, you will get a
+you run either program with the \lstinline{--help} option, you will get a
short explanation to the command line options:
\begin{verbatim}
user at server:~$ palogist --help
@@ -503,11 +504,11 @@
However, for a simple run only three options are mandatory, which
specify the necessary files needed to run the regression analysis.
-These options are \texttt{--dose} (or \texttt{-d}), specifying the
+These options are \lstinline{--dose} (or \lstinline{-d}), specifying the
genomic predictor/MLDOSE file described in section \ref{ssec:dosein};
-\texttt{--pheno} (or \texttt{-p}), specifying the phenotypic data file
-described in section \ref{ssec:phenoin}; and \texttt{--info} (or
-\texttt{-i}), specifying the SNP information file described in section
+\lstinline{--pheno} (or \lstinline{-p}), specifying the phenotypic data file
+described in section \ref{ssec:phenoin}; and \lstinline{--info} (or
+\lstinline{-i}), specifying the SNP information file described in section
\ref{ssec:infoin}.
If you change to the \texttt{examples} directory you can run
@@ -533,7 +534,7 @@
a better overview of the analysis options.
To run an analysis with MLPROB files, you need specify the MLPROB file
-with the \texttt{-d} option and also specify that there are two
+with the \lstinline{-d} option and also specify that there are two
genetic predictors per SNP, e.g.~you can run a linear model with
\begin{verbatim}
palinear -p height.txt -d gtdata/test.mlprob -i gtdata/test.mlinfo \
@@ -551,29 +552,29 @@
\end{verbatim}
\subsection{Advanced analysis options}
-The option \texttt{--interaction} allows you to include interaction
+The option \lstinline{--interaction} allows you to include interaction
between SNPs and any covariate. If for example your model is
\begin{equation*}
\textrm{trait} \sim \textrm{sex} + \textrm{age} + \textrm{SNP},
\end{equation*}
-running the program with the option \texttt{--interaction=2} will model
+running the program with the option \lstinline{--interaction=2} will model
\begin{equation*}
\textrm{trait} \sim \textrm{sex} + \textrm{age} + \textrm{SNP} +
\textrm{age} \times \mathrm{SNP}.
\end{equation*}
-The option \texttt{--robust} allows you to compute so-called
+The option \lstinline{--robust} allows you to compute so-called
``robust'' (a.k.a.~``sandwich'', a.k.a.~Hubert-White) standard errors
(cf.~section \ref{sec:methodology} ``Methodology'' for details).
-With the option \texttt{--mmscore} a score test for association
+With the option \lstinline{--mmscore} a score test for association
between a trait and genetic polymorphisms in samples of related
individuals is performed. For quantitative traits (\texttt{palinear})
a file with the inverse of the variance-covariance matrix goes as input
-parameter with that option, e.g.~\texttt{--mmscore <filename>}. The
+parameter with that option, e.g.~\lstinline{--mmscore <filename>}. The
file has to contain the first column with id names exactly like in
phenotype file, BUT OMITTING people with no measured phenotype. The
rest is a matrix. The phenotype file in case of using the
-\texttt{--mmscore} argument may contain any amount of covariates (this
+\lstinline{--mmscore} argument may contain any amount of covariates (this
is different from previous versions). The first column contains id
names, the second the trait. The others are covariates.
@@ -591,18 +592,27 @@
An example of how a polygenic object estimated by \GA{} can be used
with ProbABEL is provided in \texttt{examples/mmscore.R}
-Though technically \texttt{--mmscore} allows for inclusion of multiple
+Though technically \lstinline{--mmscore} allows for inclusion of multiple
covariates, these should be kept to minimum as this is a score test. We
suggest that any covariates explaining an essential proportion of
variance should be fit as part of \GA{}'s
\texttt{polygenic} procedure.
-%Option \texttt{--interaction\_only} is like \texttt{--interaction} but does not
-%include in the model the main effect of the covariate, which is acting in
-%interaction with SNP. This option is useful when running \texttt{--mmscore},
-%in which case the main effect should normally estimated in the polygenic
-%model and only the interaction term in the \PA{} analysis.
+The option \lstinline{--interaction_only} is like \lstinline{--interaction}
+but does not include the main effect of the covariate which is acting
+in interaction with SNP into the model. For example, if \texttt{age}
+is the second covariate in the phenotype file, using the option
+\lstinline{--interaction_only=2} leads to the following model:
+\begin{equation*}
+ \textrm{trait} \sim \textrm{sex} + \textrm{SNP} +
+ \textrm{age} \times \mathrm{SNP}.
+\end{equation*}
+This option is useful when
+running \lstinline{--mmscore}, in which case the main effect should
+normally estimated in the polygenic model and only the interaction
+term in the \PA{} analysis.
+
\subsection{Running multiple analyses at once: \texttt{probabel}}
The \texttt{bin/probabel} script is a handy wraper for the \PA{}
functions. To start using it the configuration file
@@ -663,8 +673,8 @@
frequency of the predictor allele (\texttt{A1}) in subjects with complete
phenotypic data.
-If the \texttt{--chrom} option was used, in the next column you will
-find the value specified by this option. If \texttt{--map} option was
+If the \lstinline{--chrom} option was used, in the next column you will
+find the value specified by this option. If \lstinline{--map} option was
used, in the subsequent column you will find map location taken from
the map-file. The subsequent columns provide coefficients of
regression of the phenotype onto the genotype ($\beta$), corresponding
@@ -672,7 +682,7 @@
on the likelihood ratio test. Note that for the additive, recessive,
dominant and overdominant genetic models this is a $\chi^2$ of 1
degree of freedom (df), whereas for the genotypic model this is a
-$\chi^2$ of 2df. If the \texttt{--mmscore} option is used, the
+$\chi^2$ of 2df. If the \lstinline{--mmscore} option is used, the
$\chi^2$ values are calculated from the Wald statistic
(1df)\footnote{For the genotypic (2df) model the $\chi^2$ values can't
be simply calculated using the Wald statistic, consequently the
@@ -703,7 +713,7 @@
format (functions: \texttt{mach2databel()} and
\texttt{impute2databel()}, respectively).
-When the \texttt{--mmscore} option is used, the analysis takes
+When the \lstinline{--mmscore} option is used, the analysis takes
more time.
\section{Methodology}
More information about the Genabel-commits
mailing list