From: Lasso regularization for left-censored Gaussian outcome and high-dimensional predictors
Number of HIV genotypic mutations | Gaussian BJ | NonPar BJ | LOD | LOD/2 |
present in real data study | Â | Â | Â | Â |
5 in RTG probably associated | 3 (60%) | 3 (60%) | 1 (20%) | 1 (20%) |
with abacavir resistance | Â | Â | Â | Â |
13 in RTG probably associated | 7 (54%) | 4 (31%) | 4 (31%) | 4 (31%) |
with multi-NRTI resistance | Â | Â | Â | Â |
6 in RTG probably associated | 3 (50%) | 2 (33%) | 1 (17%) | 1 (17%) |
with NRTI resistance (other than abacavir) | Â | Â | Â | Â |
30 in RTG probably associated | 12 (40%) | 11 (37%) | 8 (27%) | 7 (23%) |
with NNRTI resistance | Â | Â | Â | Â |
67 in PG probably associated | 40 (60%) | 22 (33%) | 15 (22%) | 14 (21%) |
with PI resistance | Â | Â | Â | Â |
121 Total | 65 (54%) | 42 (35%) | 29 (24%) | 27 (22%) |