We describe our statistical system for promoter recognition in genomic DNA
with which we took part in the Genome Annotation Assessment Project (GASP1)
. We applied two versions of the system: the first uses a region-based appr
oach toward transcription start site identification, namely, interpolated M
arkov chains; the second was a hybrid approach combining regions and signal
s within a stochastic segment model. We compare the results of both version
s with each other and examine how well the application on a genomic scale c
ompares with the results we previously obtained on smaller data sets.