Skip to main content

Table 4 PPB prediction results of cyclic peptide drugs for sparse modeling by ELS and FBS compared to the baseline results. These situations are shown in a part of Fig. 1(b). The values with asterisk represent the best prediction performance in each evaluation criterion, and ridge-CP-LOO and lasso-CP-LOO lines are reproduced from Table 3

From: Computational prediction of plasma protein binding of cyclic peptides from small molecule experimental data using sparse modeling techniques

Method

Training set

Test set

RMSE (fb)

MAE (fb)

R (ln Ka)

ridge-SM

(baseline #1)

Small molecules (training data)

Cyclic peptide drugs (# = 24)

0.528

0.442

0.120

lasso-SM

(baseline #2)

Small molecules (training data)

Cyclic peptide drugs (# = 24)

0.321

0.251

0.444

ELS

Small molecules (training data)

Cyclic peptide drugs (# = 24)

0.272*

0.216*

0.464*

FBS

Small molecules (training data)

Cyclic peptide drugs (# = 24)

0.381

0.288

0.270

ridge-CP-LOO

(baseline #3)

Cyclic peptide drugs (LOOCV)

Cyclic peptide drugs (LOOCV)

0.338

0.244

0.418

lasso-CP-LOO

(baseline #4)

Cyclic peptide drugs (LOOCV)

Cyclic peptide drugs (LOOCV)

0.358

0.286

0.289