

PFRMAT SS
TARGET T0086
AUTHOR 5287-1010-7667
METHOD We took 20 x (M=164) log-odds substitution matrix from psi-blast
METHOD and used this in place of 164 amino acid sequence. 
METHOD Then we took a window of 8 amino acids on either
METHOD side of a given amino acid.  This gave us 340 attributes for a
METHOD single amino acid with a structure prediction. This data was
METHOD formatted to be compatible for learning with C4.5 release 8.
METHOD 3.6 million examples were generated after windowing. 2/3rd of the
METHOD windowed data was partitioned disjointly into 16 partitions and
METHOD decision trees using C4.5 were created on each of the partitions.
METHOD The target data, T0086 was windowed in the same way as the training
METHOD data. Each learned decision tree then attempted a classification of
METHOD protein sequences in the target data and the votes were
METHOD accumulated. The percentage votes translated into the confidence
METHOD level. The class outputs were smoothed using a window of size 5, with
METHOD the prediction changed if the neighbors and the predicted AA had higher 
METHOD average confidence levels for a different structure. The average 
METHOD confidence level of each window is given to the predicted structure.
METHOD The relevant pubs and references are at 
METHOD http://morden.csee.usf.edu/~chawla for viewing.
MODEL 1
S C 0.69
H C 0.69
P C 0.65
A C 0.60
L C 0.51
T H 0.51
Q H 0.57
L H 0.61
R H 0.64
A H 0.66
L H 0.59
R H 0.57
Y H 0.54
C H 0.47
K C 0.52
E C 0.52
I C 0.56
P C 0.59
A C 0.66
L C 0.64
D C 0.64
P C 0.59
Q C 0.49
L H 0.55
L H 0.64
D H 0.65
W H 0.57
L H 0.54
L H 0.50
L H 0.47
E H 0.46
D H 0.46
S H 0.44
M H 0.47
T H 0.49
K H 0.54
R H 0.56
F H 0.56
E H 0.47
Q C 0.54
Q C 0.62
G C 0.57
K C 0.50
T E 0.45
V E 0.60
S E 0.61
V E 0.54
T E 0.46
M E 0.40
I H 0.45
R H 0.52
E H 0.46
G H 0.47
F H 0.46
V C 0.52
E C 0.59
Q C 0.57
N C 0.56
E C 0.62
I C 0.54
P C 0.49
E C 0.55
E C 0.55
L C 0.59
P C 0.69
L C 0.75
L C 0.66
P C 0.60
K C 0.55
E C 0.43
S H 0.41
R H 0.40
Y H 0.40
W H 0.46
L H 0.47
R H 0.54
E H 0.57
I H 0.55
L H 0.55
L H 0.54
C H 0.45
A C 0.52
D C 0.69
G C 0.76
E C 0.70
P C 0.60
W C 0.46
L C 0.38
A C 0.34
G C 0.35
R C 0.40
T C 0.49
V C 0.49
V C 0.50
P C 0.49
V C 0.46
S C 0.41
T C 0.41
L C 0.47
S C 0.54
G C 0.59
P C 0.55
E H 0.46
L H 0.62
A H 0.70
L H 0.75
Q H 0.65
K H 0.49
L C 0.55
G C 0.66
K C 0.75
T C 0.75
P C 0.75
L C 0.79
G C 0.74
R C 0.65
Y C 0.62
L C 0.62
F C 0.57
T C 0.65
S C 0.73
S C 0.70
T C 0.61
L C 0.55
T H 0.43
R H 0.45
D H 0.40
F E 0.43
I E 0.46
E E 0.51
I E 0.47
G C 0.39
R C 0.50
D C 0.59
A C 0.60
G C 0.56
L C 0.52
W C 0.46
G H 0.52
R H 0.60
R H 0.64
S H 0.65
R H 0.68
L H 0.64
R H 0.55
L H 0.45
S C 0.57
G C 0.61
K C 0.64
P C 0.55
L C 0.43
L E 0.39
L H 0.35
T H 0.44
E H 0.43
L C 0.47
F C 0.62
L C 0.69
P C 0.80
A C 0.84
S C 0.79
P C 0.81
L C 0.50
Y C 1.00
END


