Amplification reactions generated a fragment of the HIV genome from nucleotide 2057 to 3623, numbering according to the HXB2 (K03455) genome. The resultant PCR products were diluted 1:20 using nuclease-free water. Allele-specific PCR (AS-PCR) reactions for drug-resistant point mutations K103N, Y181C and M184V were performed, using previously published oligonucleotides [8]. Real-time PCR conditions were modified to accommodate use of the Taqman Fast Universal Master Mix (Applied Biosystems, Warrington, UK). For the M184V minority assay, we used a modified protocol PLX4032 with 55% M184V-F1, 30% M184V-F2 and 15% M184V-F3 type-specific

primers. Reactions were cycled using an ABI 7500 Fast Taqman instrument (Applied Biosystems), with an extension time of 35 s, and a reaction volume of 20 μL. Control reactions containing 1% mutant were included with each minority PCR run to provide a sensitivity cut-off point. Control reactions were generated by mixing plasmid DNA containing a subtype B reference sequence, with a second plasmid containing the same sequence with resistance point mutations. These plasmid mixtures were used to generate PCR fragments, akin to targets in

patients’ specimens, and were run alongside GSK-3 activation study specimens. All assay control reagents were identical to those used by Johnson et al. for their original technical validation investigations [8]. However, in contrast to the published methods, we included a 1% mutant control in each minority assay run. The ΔCt of these

reactions provided the sensitivity cut-off experimentally determined for each assay run. Patient-derived material with a ΔCt equal to or less than that recorded for the 1% control was scored as having minority drug resistance. All three assays underwent technical validation in house by triplicate testing of samples for reproducibility and precision, linearity of minority titrations and by testing of samples with known mixed nucleotides at relevant drug resistance codons. We also performed DNA sequencing on all patient-derived pol gene sequences. TDR was defined using mutations according Resveratrol to a published World Health Organization (WHO) list [13]. Statistical analyses were performed using McNemar’s test for paired data to compare AS-PCR methods vs. standard genotyping. Using HIV genotyping by population DNA sequencing, the K103N mutation was detected in 10 of 165 samples [6.1%; 95% confidence interval (CI) 2.9–10.9%]. Using the minority species assays, we found K103N in 12 of 165 samples (7.3%; 95% CI 3.8–12.4%). Thus, the minority-specific method increased the rate of detection of K103N by 20%; however, this was not statistically significant (P=0.5; 95% CI 0–54%).