Next: Initial state probabilities
Up: Gene Finding
Previous: A Probabilistic Model of
GENSCAN
GENSCAN, a computer program for gene identification, uses this model.
In this section we shall consider how GENSCAN determines various parameters of the model to
get meaningful results. The program uses a training set of 238
multi-exon genes and 142 single-exon genes. These are completely
sequenced genes from GenBank. On whole, the training set
consists of about 2.5 million base pairs .
Itshack Pe`er
1999-02-03