
PFRMAT SS
TARGET T0114
AUTHOR 9521-1365-5893
METHOD Program Name: PSIPRED
METHOD Synopsis: Neural network prediction based on PSIBLAST output
METHOD Five neural networks are used to predict secondary structure directly
METHOD from the profiles generated by PSIBLAST. Networks 1-4 predict
METHOD secondary structure, the 5th network filters the output of the first
METHOD level networks. The first level networks have 17*21 inputs and 75 hidden
METHOD units, the second level network has 17*3 inputs and 55 hidden units.
METHOD The networks have been trained on a set of 2937 proteins with
METHOD early-stopped training. 10% of the training data was used to detect
METHOD convergence.
METHOD The target sequence was scanned against a non-redundant databank of
METHOD filtered protein sequences (> 500,000 sequences), with 3 PSIBLAST
METHOD iterations.
MODEL  1
M C 0.96
I C 0.54
N C 0.40
R C 0.39
T C 0.40
D C 0.57
C C 0.74
N C 0.77
E C 0.78
N C 0.65
S C 0.52
Y E 0.58
L E 0.72
E E 0.75
I E 0.45
H E 0.14
N C 0.44
N C 0.66
E C 0.79
G C 0.74
R C 0.41
D C 0.08
T E 0.48
L E 0.63
C E 0.55
F E 0.34
A E 0.14
N C 0.42
A C 0.65
G C 0.69
T C 0.73
M E 0.04
P E 0.44
V E 0.69
A E 0.80
I E 0.81
Y E 0.65
G E 0.57
V E 0.60
N E 0.56
W E 0.50
V E 0.43
E C 0.12
S C 0.51
G C 0.68
N C 0.38
N C 0.07
V E 0.72
V E 0.84
T E 0.89
L E 0.90
Q E 0.80
F E 0.48
Q E 0.16
R C 0.32
N C 0.60
L C 0.76
S C 0.82
D C 0.72
P C 0.59
R C 0.42
L C 0.27
E C 0.12
T E 0.52
I E 0.63
T E 0.50
L E 0.26
Q E 0.03
K C 0.01
W C 0.38
G C 0.65
S C 0.78
W C 0.84
N C 0.61
P C 0.38
G C 0.27
H C 0.16
I H 0.01
H H 0.08
E E 0.10
I E 0.20
L E 0.30
S E 0.25
I E 0.22
R E 0.14
I E 0.05
Y C 0.86
END

