deletions | additions       diff --git a/subsection_Several_popular_methods_begin__.tex b/subsection_Several_popular_methods_begin__.tex  index 99e5765..4d62845 100644  --- a/subsection_Several_popular_methods_begin__.tex  +++ b/subsection_Several_popular_methods_begin__.tex 
...
\begin{itemize}  \item For each method on each dataset, we computed the average distance of the peak association signals from the risk region.  \item Under each method, we then plotted the empirical cumulative distribution function (ECDF) of the average peak signals for each dataset.  \item ECDF at point $x$ is the proportion of the samples with average distance less than or equal to $x$.  \item Therefore, a method with higher ECDF at a smaller value of $x$, has higher performance in localizing the signal.  \end{itemize}  \item Detection: For a given simulated dataset and a given method, we used max. statistics across all the SNV as our global test statistics and determined its null distribution by permuting the case-control labels.   \begin{itemize} 
...
\item However, test statistics are not comparable since they are not on the same scale.  \item To make these score distributions comparable, we converted scores to their corresponding empirical p values.  \item We defined these p values as the proportion of observed values under the null model that are greater than or equal to the observed value under the alternative model.  \item We then compared the distribution of empirical p values for the different test statistics by plotting empirical cumulative ditribution function (ECDF) ECDF  under each method. \end{itemize}  \end{itemize}