Clinically relevant Lys103Asn (K103N) mutant frequently observed in HIV-1 reverse transcriptase (RT) confers drug resistance. To obtain useful structural information necessary for targeted-inhibitor design, molecular docking combined with 3D-QSAR CoMFA and CoMSIA was applied to a set of 53 structurally diverse HIV-RT inhibitors. Two strategies were applied to generate 3D-QSAR models. The first strategy is the flexibility-based molecular alignment (FMA), similar to receptor-based alignment, which samples the biological space of K103N mutant HIV-RT. FMA was conducted by docking the compounds to four structural data of mutant HIV-RT with PDB codes: 1SV5, 2IC3, 1FKP and 1FKO, which are co-crystallized according to NNRTI inhibitors such as etravirine, HBY-097, nevirapine, and efavirenz. The best superposition of the compounds to the active site of 1FKP structure suggests specific inhibition of nevirapine-resistance. The second strategy is the dataset division which employs the principal component analysis (PCA) to classify the dataset into training and test sets that yields statistically significant and robust models. The PCA design selection tool by the most descriptive compounds (MDC) outperforms the largest minimum distance (LMD) for the present dataset. Overall, the results demonstrated the feasibility of the two strategies to the present case and hold a promise for its general applicability to future QSAR studies. The generated models are predictive based on reproducible values of the predicted compared with experimental activities. Further, the complementary analysis of contour maps to the mutant HIV-RT binding site suggested the anchor points for binding affinity. The present study introduced the concept 'clamp-flex' for the rational design of targeted-inhibitor to overcome the K103N pan-class resistance mutation. The predictive models offer new insights into binding modes involving the hydrophobicity and flexibility of the active site.