ubic.erminej.analysis

Class GeneSetResamplingBackgroundDistributionGenerator

java.lang.Object

ubic.erminej.analysis.AbstractLongTask

ubic.erminej.analysis.AbstractResamplingGeneSetScore

ubic.erminej.analysis.GeneSetResamplingBackgroundDistributionGenerator

All Implemented Interfaces:

NullDistributionGenerator

public class GeneSetResamplingBackgroundDistributionGenerator
extends AbstractResamplingGeneSetScore

Calculates a background distribution for class scores derived from randomly selected individual gene scores. This is used for either the "GSR" method or the precision-recall method.

Since:

Created 09/02/02.

Version:

$Id: GeneSetResamplingBackgroundDistributionGenerator.java,v 1.9 2012/05/27 02:54:30 paul Exp $

Author:

Shahmil Merchant, Paul Pavlidis

Field Summary

Fields inherited from class ubic.erminej.analysis.AbstractResamplingGeneSetScore

classMaxSize, classMinSize, hist, histogramMax, histogramMin, log, NORMAL_APPROX_SAMPLE_FREQUENCY, numClasses, numRuns, SPEDUPSIZEEXTRASTEP, SPEEDUPSIZECUT, TOLERANCE, useNormalApprox, useSpeedUp

Constructor Summary

Constructors

Constructor and Description
GeneSetResamplingBackgroundDistributionGenerator(SettingsHolder settings, Map<Gene,Double> geneToScoreMap)

Method Summary

Methods

Modifier and Type	Method and Description
protected double	averagePrecision(Collection<Gene> genesInSet)
double	computeRawScore(double[] genevalues, Collection<Gene> genesInSet) Basic method to calculate the raw score for a gene set, given an array of the gene scores for items in the class.
Histogram	generateNullDistribution(StatusViewer m) Used for methods which require randomly sampling classes to generate a null distribution of scores based on gene-by-gene scores.
Histogram	generateNulldistribution(StatusViewer m, long randomSeed) Generate a null distribution, using a selected random seed.
void	setHistogramRange() Figure out the limits that should be used for the histogram.
void	setQuantile(int value)
void	setRandomSeed(long randomSeed)

Methods inherited from class ubic.erminej.analysis.AbstractResamplingGeneSetScore

getHist, setClassMaxSize, setUseNormalApprox, setUseSpeedUp

Methods inherited from class ubic.erminej.analysis.AbstractLongTask

ifInterruptedStop

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

GeneSetResamplingBackgroundDistributionGenerator

public GeneSetResamplingBackgroundDistributionGenerator(SettingsHolder settings,
                                                Map<Gene,Double> geneToScoreMap)

Parameters:

settings -

geneScores - Should already be multifunctionality corrected if desired.

Method Detail

computeRawScore

public double computeRawScore(double[] genevalues,
                     Collection<Gene> genesInSet)

Basic method to calculate the raw score for a gene set, given an array of the gene scores for items in the class. Note that speed here is important. In the prototypical GSR method, the score is the mean of the values for the gene. For precision-recall, it is the average precision.

Parameters:

genevalues - double[] raw scores for the items in the class.

genesInSet -

Returns:

double

See Also:

for choices of methods.

generateNulldistribution

public Histogram generateNulldistribution(StatusViewer m,
                                 long randomSeed)

Generate a null distribution, using a selected random seed.

Parameters:

m -

randomSeed -

Returns:

generateNullDistribution

public Histogram generateNullDistribution(StatusViewer m)

Used for methods which require randomly sampling classes to generate a null distribution of scores based on gene-by-gene scores.

Parameters:

m -

Returns:

A histogram object containing a cdf that can be used to generate pvalues.

setHistogramRange

public void setHistogramRange()

Figure out the limits that should be used for the histogram.

setQuantile

public void setQuantile(int value)

Parameters:

value - int

setRandomSeed

public void setRandomSeed(long randomSeed)

averagePrecision

protected double averagePrecision(Collection<Gene> genesInSet)

Parameters:

genesInSet -

Returns: