Biomedical Engineering Reference
In-Depth Information
transporters and further classified into six families. Li et al. [
84
]utilizednearest
neighbor algorithm to distinguish 484 transporter families and reported a fivefold
cross validated accuracy of 72.3%.
4.5 Channels and Pores
In TCDB, channels and pores are grouped in the same class to include the transport
systems that catalyze facilitated diffusion (by an energy-independent process) by
passage through a transmembrane aqueous pore or channel without evidence for a
carrier-mediated mechanism. However, channels have
a
-helical conformation,
whereas pores have
-strands in their membrane spanning segments.
We have tested our method to discriminate channels and pores and the results
obtained with amino acid composition are shown in Table
5
. We found that most of
the machine-learning methods discriminated the channels and pores with the
accuracy in the range of 88-92%. The neural network and support vector machine
showed the highest accuracy of 92.4%. The sensitivity and specificity are 93 and
92%, respectively using neural network. The achievement of high accuracy might
be due to the difference in amino acid residues in the membrane spanning regions of
a
b
-helical membrane proteins are
dominated with the stretches of hydrophobic residues, whereas the polar and
charged resides are intervened in the membrane spanning segments of
-helical and
b
-barrel membrane proteins. The
a
b
-barrel
membrane proteins.
Table 5 Discrimination of channels and pores using different machine learning approaches [
78
]
Fivefold cross-validation
Sensitivity (%)
Method
Specificity (%)
Accuracy (%)
Bayesnet
94.1
81.4
88.9
Naive Bayes
92.5
88.4
90.8
Logistic function
92.0
89.1
90.8
Neural network
93.0
91.5
92.4
RBF network
92.5
88.4
90.8
Support vector machines
95.2
88.4
92.4
k-nearest neighbor
89.8
86.8
88.6
Bagging meta learning
89.8
83.7
87.3
Classification via Regression
88.2
85.3
87.0
Decision tree J4.8
86.1
78.3
82.9
NBTree
90.9
83.7
88.0
Partial decision tree
87.2
79.1
83.9
Sensitivity
(TP+TN)/(TP+TN+FP+FN)
TP, FP, TN and FN refer to the number of true positives, false positives, true negatives and false
negatives, respectively
¼
TP/(TP+FN); Specificity
¼
TP/(TP+FP); Accuracy
¼
