Alu: /* Details */

2011-10-26T13:04:17Z

Details

← Older revision		Revision as of 15:04, 26 October 2011
Line 28:		Line 28:
	=== Details ===		=== Details ===

	~~[[SignificaneTest]]~~ uses the R method p.adjust to compute adjusted p-values. Therefore, it provides the methods listed under '''padj.method'''. ''None'' will put the same as the unadjusted p-values.		The significance test uses the R method p.adjust to compute adjusted p-values. Therefore, it provides the methods listed under '''padj.method'''. ''None'' will put the same as the unadjusted p-values.

	=== References ===		=== References ===

Admin: 2 revisions

2011-10-26T06:16:11Z

2 revisions

← Older revision	Revision as of 08:16, 26 October 2011
(No difference)

imported>MichaelDondrup at 15:56, 4 April 2006

2006-04-04T15:56:16Z

← Older revision		Revision as of 17:56, 4 April 2006
Line 1:		Line 1:
	__NOTOC__		__NOTOC__
			<!-- ## page was renamed from EMMAWiki/Web Documentation/AnalysisFunctions -->
	<!-- ##language:en -->		<!-- ##language:en -->
	== Significance Test ==		== Significance Test ==

Unknown user at 00:00, 1 January 1970

1970-01-01T00:00:00Z

New page

__NOTOC__

== Significance Test ==

=== Description ===
Test for significantly up- or down-regulated genes, by using either a t-test or wilcox-test. This method allows to compute adjusted p-values for the test. In addition, some more simple statistics are computed.

=== Input ===
Works for one or many normalized datasets.

=== Output ===

A table of data containing the following values:

* '''Statistic:''' The test statistik (either t or w value from the test)
* '''p-value:''' The p-value of the test statistik for this gene
* '''adjusted p-value:''' The ''adjusted'' p-value according to the adjustment method set
* '''M1-mean:''' The mean of the log-ratio over replicates
* ...

=== Parameters ===

* '''testtype''' Type of test to use [t.test | wilcox test]
* '''padj.method''' Chose method for adjusting the p.values [none | hommel | hochberg | fdr | bonferroni]
* '''two.sample''' Do a two-sample test? Otherwise one-sample test.

=== Details ===

[[SignificaneTest]] uses the R method p.adjust to compute adjusted p-values. Therefore, it provides the methods listed under '''padj.method'''. ''None'' will put the same as the unadjusted p-values.

=== References ===

==== Adjusted p-values (correct for multiple testing) ====

(the following references are taken from the R help page)

* Benjamini, Y., and Hochberg, Y. (1995). Controlling the false
discovery rate: a practical and powerful approach to multiple
testing. _Journal of the Royal Statistical Society Series_ B,
* 57*, 289-300.
* Holm, S. (1979). A simple sequentially rejective multiple test
procedure. _Scandinavian Journal of Statistics_, *6*, 65-70.
* Hommel, G. (1988). A stagewise rejective multiple test procedure
based on a modified Bonferroni test. _Biometrika_, *75*, 383-386
*
* Hochberg, Y. (1988). A sharper Bonferroni procedure for multiple
tests of significance. _Biometrika_, *75*, 800-803.
* Shaffer, J. P. (1995). Multiple hypothesis testing. _Annual Review
of Psychology_, *46*, 561-576. (An excellent review of the area.)
* Sarkar, S. (1998). Some probability inequalities for ordered MTP 2
random variables: a proof of Simes conjecture. _Annals of
Statistics_, *26*, 494-504.
* Sarkar, S., and Chang, C. K. (1997). Simes' method for multiple
hypothesis testing with positively dependent test statistics.
_Journal of the American Statistical Association_, *92*,
1601-1608.

* Wright, S. P. (1992). Adjusted P-values for simultaneous
inference. _Biometrics_, *48*, 1005-1013. (Explains the adjusted
P-value approach.)

EMMAWiki/WebDocumentation/AnalysisFunctions/SignificanceTest - Revision history

Alu: /* Details */

Admin: 2 revisions

imported>MichaelDondrup at 15:56, 4 April 2006

Unknown user at 00:00, 1 January 1970