SUPPLIMENTARY MATERIAL

There are following six suplimentary materials


Table S1:Performance of SVM based BhairPred algorithm trained on Thronton's dataset

APPROACH PERCENT COVERAGE PERCENT PROBABILITY ACCURACY MCC
Qo(H) Qo(Nh) Qp(H) Qp(nh)
AA 68.1 63.2 69.4 62.0 65.9 0.31
MSA (1) 77.3 65.7 73.4 70.7 72.1 0.44
AA+ACC_O (2) 78.4 57.06 69 68.12 68.6 0.36
AA+SS_O (3) 65.6 75.1 76.3 64.3 69.9 0.41
HYBRID (1+2+3) 78.8 70.6 76.6 73.3 75.1 0.49
AA+ACC_P (4) 70.8 59.4 68.0 62.8 65.7 0.30
AA+SS_P (5) 60.76 74.8 74.7 61.0 67.1 0.36
HYBRID (1+4+5) 81.5 58.3 70.5 72.4 71.1 0.41

GO TO TOP

Table S2:Performance of ANN based BhairPred algorithm trained on Thronton's dataset

APPROACH Qo(H) Qo(Nh) ACCURACY Qp(H) Qp(nh) Sh Snh MCC
AA 45.7 71.9 58.1 69.1 51.2 5.9 16.2 0.19
MSA 45.8 75.9 59.3 70.0 53.3 6.9 20 0.23
AA+SS_O 67.7 75.8 71.3 77.4 65.9 14.7 9.9 0.43
AA+SS_P 55.6 75.8 62.7 70.4 56.8 8.8 11 0.27
AA+ACC_O 55.8 73.5 63.5 72.2 57.7 9.5 13.1 0.28
AA+ACC_P 46.6 68.5 57.1 65.1 51.2 4.8 13.1 0.16
Seq-Str Newtowrk (SS_P) 55.1 74.8 63.9 73.3 57.8 9.7 14.7 0.31
Seq-Str Newtowrk (SS_O) 64.9 74.3 69.1 75.7 63.4 13.1 9.5 0.39

AA:Amino Acid Sequence: MSA:Multiple Sequence Alignment; ACC_O:Observed Accessibility; ACC_P:Predicted Accessibility; SS_O:Observed Secondary Structure; SS_P:Predicted Secondary Structure

GO TO TOP

Figure S2:Training scheme of ANN

GO TO TOP

Figure S3:Training scheme of SVM

GO TO TOP

Figure S4:Result of consensus prediction of BhairPred on 2880 proteins dataset

GO TO TOP

Figure S5:Result of consensus prediction of BhairPred on 2880 proteins dataset

GO TO TOP

Figure S6:Submission form of BhairPred server

GO TO TOP

Figure S7:Prediction result of BhairPred server

GO TO TOP

Sensitivity:It is the total correct prediction of beta hairpin among all the hairpins

GO TO TOP

Specificity:It is the total correct prediction of of non-beta hairpin among all non hairpins

GO TO TOP

Accuracy:It is the total fraction of correct prediction of of beta and non-beta hairpins divided by total number of predictions

GO TO TOP

Mathew's Correlation Coefficient (MCC):It is a measure to judge the quality of prediciton

GO TO TOP

Qp(H):It is probability of positive or hairpin prediciton

GO TO TOP

Qp(Nh):It is probability of negative or non-hairpin prediciton

GO TO TOP

P:True Positive: N:True Negative; O:False Positive; U:False Negative;