Benchmarking Machine Learning Models for HIV-1 Protease Inhibitor Resistance Prediction: Impact of Data Set Construction and Feature Representation.

Statistical analysis showed a strong association between sequence-based clusters and resistance levels, with both one-way ANOVA ( F 4 , 859 = 277.49, p = 4.59 × 10 –153 ) and Kruskal–Wallis tests ( H = 508.74, p = 8.63 × 10 –109 ) confirming highly significant differences in mean log 10 (fold change) across groups ( Figure A and Table S3 ).

also in 132,142 other papers