Skip to main content
. 2014 Apr 14;4:4668. doi: 10.1038/srep04668

Table 1. Pearson correlation and root mean square error values obtained for each pancreatic cell line by their respective QSAR models.

S. No. Cell Line ND* D1 R R2 RMSE D2 R R2 RMSE
1. AsPC-1 92 45 0.84 0.70 1.6 23 0.81 0.66 1.78
2. BxPC-3 92 47 0.82 0.67 1.82 19 0.82 0.67 1.83
3. CAPAN-1 96 42 0.83 0.69 1.61 21 0.84 0.71 1.58
4. Capan-2 90 54 0.86 0.73 1.47 22 0.82 0.67 1.63
5. HPAF-II 95 42 0.86 0.74 1.45 22 0.85 0.73 1.48
6. HuP-T3 96 54 0.85 0.73 1.57 22 0.84 0.71 1.63
7. HuP-T4 92 41 0.79 0.61 2.00 24 0.83 0.69 1.81
8. KP-4 88 62 0.82 0.68 1.94 22 0.82 0.67 2.04
9. MIA-PaCa-2 96 45 0.85 0.71 1.81 23 0.86 0.73 1.76
10. MZ1-PC 132 136 0.79 0.61 1.94 29 0.73 0.53 2.16
11. PANC-03-27 91 44 0.8 0.64 2.00 20 0.82 0.67 1.91
12. PANC-08-13 92 50 0.81 0.65 1.75 22 0.82 0.68 1.74
13. PANC-10-05 95 41 0.81 0.65 1.92 24 0.85 0.73 1.71
14. PSN1 129 116 0.64 0.42 2.68 31 0.63 0.39 2.81
15. SW1990 94 38 0.77 0.56 1.89 23 0.80 0.63 1.83
16. YAPC 86 65 0.89 0.79 1.24 21 0.86 0.74 1.45

*ND: Number of drugs used to develop QSAR models; D1: Number of descriptors selected from the CfsSubsetEval algorithm; D2: Number of descriptor selected from CfsSubsetEval followed by F-stepping; R: Pearson Correlation; RMSE: Root Mean Square Error.