qvalue

Usage:

qvalue [options] <p-values file>

Description

Takes as input a list of p-values and prints a corresponding list of q-values, computed using the method of Benjamini and Hochberg (Journal of the Royal Statistical Society B, 57:289-300, 1995).

Input

<p-values file>

The name of a tab separated file containing a column of p-values. Other columns may exist. If the name is - then the input data is read from standard input.

Output

Outputs tab-delimited columns. If the --append option is given, then the line from the input is listed first, followed by the p-value (only when --null is specified) and q-value. Otherwise it outputs the raw score (only when --null is specified). followed by the p-value and q-value. The lines are sorted by p-value.

Options

Option Parameter Description Default Behavior
General Options
--nullfile Treat the primary input file as raw scores, and begin by estimating p-values from the empirical null scores given in this file. The --header and --column arguments apply to this file as well as the primary input file. The primary input is assumed to actually contain p-values.
--good-scorehigh|​low Choose low if smaller scores are better or high if larger scores are better. Smaller scores are considered better.
--pi-zero Estimate the fraction (pi-zero) of the empirical distribution that is drawn according to the null hypothesis. This uses the method described in:
John Storey, "A direct approach to false discovery rates", Journal of the Royal Statistical Society B, 64:3, 479-498, 2002.
The estimation step is skipped and the program conservatively sets pi-zero to 100%.
--pi-zero-filefile name Specify a file to store the pi-zero estimate in. This option implies the --pi-zero option. The results of the pi-zero estimate (if calculated) are not stored in a file.
--fdr Only compute false discovery rate (FDR) estimates. The FDR values are converted into q-values.
--bootstrapsnum Number of bootstraps to perform when estimating pi-zero. The program performs 1000 bootstraps.
--headerline count Treat the first line count lines of the p-values file as the header. If the --append option is specified these lines will also be included in the output. No lines are treated as headers.
--columncolumn number Read the p-value from the specified column number. The first column is 1, the second column is 2 and so on... The p-value is read from the first column.
--append Print all the columns in the input file to standard output with additional p-value (raw scores only) and q-value columns. Print the raw score (only when raw scores are supplied), p-value and q-value in tab separated columns.
--seedseed Set the seed value for the pseudo-random number generator. The seed value is set from the system clock.