Skip to main content

Table 2 Comparison of the LASSO model and the BIOPREDsi algorithm on difference training and test sets.

From: An accurate and interpretable model for siRNA efficacy prediction

Training Set

All (249)

All human (198)

hE2 (139)

Rodent (51)

All (2,182)

0.67/0.66

0.66/0.63

0.67/0.63

0.75/0.77

All human (1,744)

0.66/0.65

0.65/0.61

0.66/0.62

0.72/0.72

Human E2s (1,229)

0.65/0.65

0.64/0.62

0.65/0.62

0.71/0.76

Rodent (438)

0.57/0.55

0.55/0.54

0.55/0.53

0.68/0.57

Random all (1,091)

0.66/0.65

0.64/0.62

0.64/0.61

0.76/0.75

Random all (727)

0.67/0.65

0.66/0.63

0.66/0.63

0.74/0.76

Random all (545)

0.61/0.62

0.61/0.60

0.61/0.60

0.65/0.70

Random all (218)

0.57/0.47

0.56/0.47

0.55/0.46

0.60/0.46

  1. This table shows the Pearson correlation coefficients between predicted and actual efficacy on training and test data sets of various sizes and composition. Each cell reports two values. The first one is the coefficient obtained by the LASSO method with the combination of the sparse representation on 21 nucleotides and the spectral representation on the first 19 nucleotides only. The second value is the coefficient reached by BIOPREDsi on the same dataset (from Huesken et al. [33].)