Table 1.
Median and 95% and 99% quantiles of each distance method for protease (PR) and reverse transcriptase (RT).
| Gene, drug-resistance mutations,a distanceb |
Median | 95% quantile |
99% quantile |
|---|---|---|---|
| PR | |||
| Unadjusted | |||
| Hamming | 1.7 | 4.7 | 6.1 |
| Mixture-weighted | 1.0 | 3.7 | 5.4 |
| Unweighted | 0.3 | 3.0 | 4.7 |
| Adjusted | |||
| Hamming | 1.3 | 3.4 | 4.7 |
| Mixture-weighted | 0.6 | 2.4 | 3.4 |
| Unweighted | 0.0 | 1.7 | 3.0 |
| RT | |||
| Unadjusted | |||
| Hamming | 1.5 | 3.6 | 4.3 |
| Mixture-weighted | 0.9 | 2.7 | 3.5 |
| Unweighted | 0.3 | 1.9 | 2.9 |
| Adjusted | |||
| Hamming | 1.3 | 3.4 | 4.7 |
| Mixture-weighted | 0.7 | 2.1 | 2.8 |
| Unweighted | 0.1 | 1.3 | 2.2 |
NOTE. Data are percentage nucleotide distance.
a
The mutation-adjusted distance excludes codons associated with drug resistance. The unadjusted distance includes all 99 PR positions and 250 RT positions.
b
Three approaches were used to handle nucleotide mixtures. The Hamming distance weighed all nucleic acid differences between 2 sequences equally. The mixture-weighted distance weighed nucleotide distances according to the extent of overlap between 2 nucleotides (e.g., A and R, which represent a mixture of A/G overlap by 50%). The unweighted distance ignored positions that had nucleotide mixtures.