Skip to main content
. 2020 Jan 22;10:1550. doi: 10.3389/fphar.2019.01550

Table 4.

Mean absolute error and mean percentage within 20% of actual dose stratified by therapeutic warfarin dose requirements (i.e., sensitives, resistant and normal) in the test set of the Caribbean Hispanic cohort.

Models Normal Sensitive Resistant
MAE (mg/week) Within 20% MAE (mg/week) Within 20% MAE (mg/week) Within 20%
RPART 4.17 (2.60-5.73) 88.00 9.83 (3.53-16.12) 37.50 14.28 (7.19-21.37) 33.33
MARS 6.98 (4.44-9.52) 68.00 11.44 (6.55-16.33)# 25.00 7.22 (2.11-12.33)# 66.67
RFR 2.91(2.18-3.64)# 100.00 7.17 (3.48-10.86)# 50.00 13.45 (7.41-19.48) 33.33
ANN 8.03 (5.01-11.05) 68.00 10.32 (1.49-19.15) 50.00 16.75 (10.87-22.63)# 0.00
SVR 5.67 (3.82-7.53)# 68.00 4.79 (1.21-8.36) 75.00 19.44 (15.83-23.05) 0.00
iBK1 3.62 (2.01-5.24) 88.00 12.81 (6.15-19.48)# 25.00 12.83 (2.48-23.18) 33.33
iBK2 3.75 (1.94-5.55) 88.00 9.33 (2.17-16.49)# 37.50 7.58 (2.33-17.50) 66.67
iBK3 3.83 (2.01-5.64) 88.00 9.33 (2.17-16.49)# 37.50 7.58 (2.33-17.50) 66.67
REPT 6.89 (4.68-9.09)# 56.00 12.30 (2.60-22.01)# 37.50 15.08 (2.67-27.50) 66.67

#best prediction model does not include common variants in VKORC1 (rs9923231), CYP2C9 (rs1799853), CYP4F2 (rs2108622) and NQO1 (rs1800566).

Data are expressed as mean (95% CI) or percentage. MAE: mean absolute error; multivariate adaptive regression splines (MARS), artificial neural networks (ANN), random forest regression (RFR), support vector regression (SVR), K-nearest neighbor for K from 1 to 3 (iBK1, iBK2, iBK3), recursive partitioning (RPART) and reduces error pruning tree classifier (REPT).