The following links load some of the above data sets and allow you to start an actual run of Pomelo II. You do need to select the options yourself; we only load the data.

Contingency tables: Fisher IxJ

Contingency table example. Analyze using "Fisher IxJ" options under Test.

t-test

Leukemia example. Analyze using a permutation t-test or a t-test with limma.

ANOVA and other linear models

Srbct example. A linear model with four classes. You must select Anova linear models test (under Test options) and select "Continue without covariables" in the second input screen.

Breast example. A linear model with three classes and additional covariates. You must select Anova linear models test (under Test options), and select "Send covariables" in the second input screen. In the third screen, select whichever covariables you want.

Survival analysis (Cox model)

An example with censored survival data. You must select Cox model (under Test options).

An example with clickable links

This is a survival data example with very little signal. By default, the results show no heatmap, but of course you can change the options and draw a heatmap. The heatmap and table both are clickable and will give you more data on the genes, based on the IDConverter databases. (These are the same data as used in the Named Genes example of signs.)

Leukemia data set of Golub et al.

This is a classical example. There is quite a bit of signal in these data and by default you obtain a heatmap which, it turns out, is a good-looking one. By default, only the first 50 genes are shown, but you can modify the options of the heatmap to show many more genes that show a strong signal.

Sending results to PaLS

Here you can examine the sets of GO terms, KEGG and Reactome pathways, and PubMed references in the set of selected genes (you choose the criteria for selection). Scroll to the bottom of the page, and click the "Send genes to PaLS", to be taken to a screen where you select the criteria for sending the genes for PaLS analysis.

Data sets

The data for the Fisher's test example are fictitious. Details on the data sets Leukemia, Srbc and Breast, including their origin and preprocessing are available from the supplementary material to the paper Gene selection and classification of microarray data using random forest. The additional covariates for the Srbct example are fictitious.

Contingency table test example

"Expression" data file and Labels data file.

Leukemia example

Expression data file and Labels data file.

Srbct example

Expression data file, Labels data file, and

Breast example

Expression data file, Labels data file, and Additional covariates data file.

Survival analysis example

Expression data file, Survival data file, and Censored indicator data file.

Download all the data sets as tar.gz or as a zip file.

The text of this document and the comments added to the output are copyright © 2005-2009 Ramón Díaz-Uriarte. You can, however, do quite a bit of things with all the text, since this document is licensed under a Creative Commons license, as shown below (click on the links to read the terms of the license).

Go to main help