File format description for {train,test}_dat.split[1-5]
=======================================================

each datum is represented by 7 lines. 
line 1: represents the class label (-1 or +1 wheras +1 mean alternative exon, -1 non alternative exon)
line 2: window around acceptor site (-200 +200)  -> 401 long
line 3: window around donor site (-200 +200)     -> 401 long
line 4 to 6: exon, upstream intron, downstream intron length features 
and in which of the three frames the of the exon stop
codons appear (for more details see paper)


e.g.

-1
ctggattaccgtcgattcggtttgtaattctga...
aatgacggcttactctgactttacacctcaaga...
1.000000 1.000000 1.000000 1.000000 ...
1.000000 1.000000 1.000000 1.000000 ...
1.000000 1.000000 1.000000 1.000000 ...
0.000000 1.000000 0.000000 0.000000 ...
