Research > Software > FDRShape

FDRShape

FDRShape is an application that performs a False Discovery Rate procedure on the p-values associated with the vertices of a shape.

Features

  • <p>The application allows the user to set the q-value(s). The user may specify the q-values with a flag or by a simple text file containing the q-values.</p>
  • <p>With the proper flag, the program will appropriately use negative p-values.</p>
  • <p>With the proper flag, the rejected p-values will be replaced by a user-specified floating point p-value.</p>

Description

FDRShape is an application that performs a False Discovery Rate procedure on the p-values associated with the vertices of a shape.

Usage

System Requirements

  • OS: any
  • Processor: any
  • Memory: Data-dependent
  • Other: Java 1.5 (or higher) runtime environment

Installation

Compilation
Portable java executable make compilation unnecessary.

Installation Instructions
Copy java executable 'jar' file to local disk.

Purpose

FDRShape takes, as input, a shape with floating point vertex attributes that represent the p-values associated with each vertex. The False Discovery Rate of the p-values is then determined, and the results are written to the floating point attributes of a shape. Both Ucf and Dx surfaces are compatible with the application.

Determination of statistical significance of a set of p-values is often excessively conservative when applied to neurological hypotheses. The False Discovery Rate is a more lenient measure of significance. This application enables a user to apply the False Discovery Rate threshold to a shape containing statistical p-values as vertex attributes.