Software

J/maanova

Current Version 0.8 (May 30th, 2006)

[ Download | Installation | Status | Screen shots | Manual (pdf) ]

J/maanova is a software for statistical analysis of MicroArray experiments. Written in Java and using R/maanova as the engine, J/maanova provides a powerful and interactive working environment for users to perform a variety of tasks.

How it works

J/maanova is a graphical user interface for R/maanova, which is an add-on package for the freely available statistical language R. While R/maanova is a powerful data analysis software package, it requires some programming skill. We developed J/maanova to help the people without programming skill to analyze their data. The software works in such a way that people do their job through the point click interface. The software translates the users' actions into R command, run the command in R, get the result back and display them. So R and R/maanova are the actual work horses. J/maanova uses a third party software Rserve as the middleman between the GUI and R.

What is R and why R

R is a programming language and environment for statistical computing. It provides a wide variety of statistical functions and is highly extensible. Plus R is open source and free of charge. Syntax wise R is very similar to S (And Splus).

Given the rich collections of statistical tools provided by R, rewriting them will be a waste of resources. So we decided to utilize them and make the development easier.

Why R/maanova

R/maanova is our free, open source package for N-dye microarray data analysis. It is implemented as an add-on package for R. R/maanova is a powerful command-line driven analysis system. Unlike many other Microarray data anlysis softwares which use log-ratio
based analysis, R/maanova takes the whole experimental design into account. ANOVA provides a general approach to the analysis of multiple factor experiments. Mixed model ANOVA provides a way to consider multiple sources of variation in the microarray
experiment. In general, R/maanova can be applied to any microarray experimental designs but it is specialized in more complicated multiple factor designs. Given an experiment like that user can ask a variety of questions and draw statistical inferences.

Why Java

The nicest feature of Java is the cross-platform ability. So J/maanova will run on a variety of operating systems like Windows, Linux and Mac OS. Java provides a rich collection
of GUI components so we can make the interface looks nice and highly interactive. For example, in J/maanova user can click a point in the volcano plot to view the gene information and its profile plot. Performance wise Java is inferior to some other languages like C++ and Visual Basic. But since the engine functions are running in R, that is not a big concern.

Author

The software and manul were written by Hao Wu and Lei Wu. Some Java codes were borrowed from TIGR's MultiExperiment Viewer (TMEV) software.

Contact

For problems/questions/suggestions, please send email to maanova@jax.org.

There is an online discussion group in google. You are welcome to join the group and ask/answer questions, make comments and suggestions, share experiences with other users.