I have released an automatic-annotation program for whole-genome
tiling-array data. The program named ARTADE and its source program is
now available at the following web-site.
ARTADE calculates gene structures (exon/intron structure), their
expression levels, and statistical significances of their expressions,
based both on the genomic sequence data and array data. ARTADE
automatically predicts unknown gene structures in the inter-genic
regions where no known gene exists. In the prediction process, ARTADE
calculates the likelihood of the structure based on a Markov model, and
thus, predicts gene splice points accurately, and overcome the
disadvantages of conventional threshold-based prediction methods.
You can use it on various platforms including Windows, Mac, Linux, 64bit
machines, because the program is written in pure C++, and does not
require any additional library such as STL. I designed its minimum
memory requirement is 512MB, where it is possible to perform a whole
chromosomal analysis in a single run. However I recommend 1GB memory for
maximizing its performance.
So far, ARTADE is applicable to Arabidopsis thaliana data only. I am
planning to release ARTADE for mouse and other species until the end of
You can browse examples of predicted genes by ARTADE on our GPS database
system at the web-site.
Tetsuro Toyoda, ph.D.
Genomic Sciences Center, RIKEN, Japan.