README file for training.tar.gz
Uta Schulze, 18-Jul-2006

The following files are in FASTA format:

training_dna.fa			(4604 DNA sequences)
training_query_noise=0.00.fa	(4604 EST/cDNA sequences)
training_query_noise=0.01.fa	(4604 EST/cDNA sequences)
training_query_noise=0.10.fa	(4604 EST/cDNA sequences)
training_query_noise=0.20.fa	(4604 EST/cDNA sequences)
training_query_noise=0.50.fa	(4604 EST/cDNA sequences)

The description lines in each FASTA file contain an identifier. In addition, the headers in the files training_dna.fa and training_query_noise=0.00.fa contain the true exon boundaries created by an alignment between a DNA sequence with its corresponding EST/cDNA (query) sequence.
They are written the following way: "exon1_start..exon1_end,exon2_start..exon2_end,exon3_start..exon3_end"

For example a DNA sequence is:
>train_dna1, 1..59,229..243,561..693
atggacgatgagggggagtttgtgttgtatctccgttcactgaccggttaatcagtgagg
tgctactcgtttttttctttgtttgagaattaatgttgatcaacaatacttaggtgaaca
tttgaattaatggcattttacccaaagtttcgaactcgaatatcctattttactggggtt
ttgcactattttccatttattgtaaattcattttatttattttttcagaaatgatatctg
caggtatgcgacttttcagcaaaattgattgtgtatattctagaaggtctctgagtagac
atacttgaaagtatacactccaggagtaactctcctcttattcaagacaaattcaaagca
atgttcatttcctacaatgttcaaatacaaacatgcgaacactattaaattataaaatct
gaagaaaaaaacgttttttttttgacaattaccaaatttatgaaaaagtaactctattag
aaatcattcaaaaaatcacatcggatataagtgtaatttgatttttttttaaactatatg
ccatattacaatattttcaggccctcgacatcgaaaatcccgaaaacgagaatcaactgg
aagtagtggaagttcagaagaagaaaaaagaccacgtactgcattcactggagatcaatt
ggaccggctcaaaactgaattccgggaaagcag

and the corresponding EST sequence:
>train_noise=0.00_query1, 1..59,60..74,75..207
atggacgatgagggggagtttgtgttgtatctccgttcactgaccggttaatcagtgaga
aatgatatctgcaggccctcgacatcgaaaatcccgaaaacgagaatcaactggaagtag
tggaagttcagaagaagaaaaaagaccacgtactgcattcactggagatcaattggaccg
gctcaaaactgaattccgggaaagcag