E born in North America and infected with subtype B HIV-1 virus. The single African subject was infected ?with a subtype C virus. Twelve subjects were ART-naive, 3 subjects were receiving ART, while 2 had received ART therapy in the past. The duration of HIV infection, was between 2 months and 20 years. Subjects had a detectable VL with the range 504 ,50,192 copies/ml. Only 3 of 17 subjects had CD4 counts #200 cells/ml. (Table 1). Plasma, PBMCs and DBS for all subjects were successfully sequenced by TPP. A total of 49,222 TPP reads were used in downstream analysis. The average oversampling rate was 290 for each nucleotide position for the 17 sets of specimens. The net TPP error rate was 0.35 as measured using pedigreed plasmid controls. Nucleotide variations identified at the 5 MBIT level were reevaluated through pairwise comparison among plasma, PBMCs and DBS, using consensus CASIN custom synthesis sequences generated with an MBIT of 20 [4,18]. The average SCR for each subject were 82.9611.9 for DBS vs. plasma, 78.9610.9 for DBS vs. PBMC and 75.31614.7 for plasma vs. PBMC. SCR were re-analyzed after subjects were stratified according to VL, ART status, CD4 counts, and duration of HIV infection. When compared to subjects with a VL of .5,000 copies/ml, subjects with a lower VL had a greater sequence Ergocalciferol web discordance between plasma and either DBS or PBMC. The mean SCR for DBS vs. plasma was 72.0 when the VL was ,5,000 copies/ml, and 88.8 when the VL was 5,000 copies/ml (p = 0.002). The plasma vs. PBMC SCR was 65.7 vs 80.6 (p = 0.042) in the same low/high viral load stratification. No significant difference in the SCR was found when comparing sequences from DBS and those from PBMC at different VL (75.3 vs 80.9 , p = 0.325) (Figure 1a). Current ART use also had a significant impact on the SCR. When stratified according to ART status, the mean SCRs in the ?off-ART group (ART-naive or prior ART) were 86.3 , 80.7 and 78.6 for the pairs of DBS vs. plasma, DBS vs. PBMC and plasma vs. PBMC respectively. In contrast, the corresponding SCRs in the on-ART groups were 67.2 , 70.5 and 59.9 respectively. The inter-group differences were statistically significant for plasma vs.DBS and plasma vs. PBMCs pairs (p = 0.007and 0.041 respectively). Although not statistically significant, the DBS vs. PBMC comparison showed a trend to greater differences in SCR when the patients were off-ART (p = 0.146) (Figure 1b). Although ART exposure appeared influences SCR, multivariate analysis demonstrated that VL level was the only independent correlate of SCR when comparing DBS with plasma. The duration of HIV infection was correlated with the SCRs found when the specimens were compared (Figure 1C). For example, a significantly higher SCR between DBS- and plasmabased HIV genotypes were observed in patients with an infection of #2 years as compared to those infected by HIV for longer periods (90.5 vs 82.8 , p = 0.044) . This finding was also seen when plasma and PBMC genotypes were compared. (SCR 84.0 vs 74.8 , p = 0.05). The same SCR trend was observed for the DBSvs. PBMC comparison although the difference was not significant (p = 0.08).Subjects and specimenAfter informed consent, 17 HIV-1 positive subjects provided an EDTA anti-coagulated blood specimen. A FACSCalibur flow cytometer (BD Biosciences, USA) was used for CD4+ cell enumeration. DBS were prepared by pipetting 75 ml/spot of whole blood onto Whatman 903H filter paper (Whatman Inc, Florham Park, USA). Each card was air-d.E born in North America and infected with subtype B HIV-1 virus. The single African subject was infected ?with a subtype C virus. Twelve subjects were ART-naive, 3 subjects were receiving ART, while 2 had received ART therapy in the past. The duration of HIV infection, was between 2 months and 20 years. Subjects had a detectable VL with the range 504 ,50,192 copies/ml. Only 3 of 17 subjects had CD4 counts #200 cells/ml. (Table 1). Plasma, PBMCs and DBS for all subjects were successfully sequenced by TPP. A total of 49,222 TPP reads were used in downstream analysis. The average oversampling rate was 290 for each nucleotide position for the 17 sets of specimens. The net TPP error rate was 0.35 as measured using pedigreed plasmid controls. Nucleotide variations identified at the 5 MBIT level were reevaluated through pairwise comparison among plasma, PBMCs and DBS, using consensus sequences generated with an MBIT of 20 [4,18]. The average SCR for each subject were 82.9611.9 for DBS vs. plasma, 78.9610.9 for DBS vs. PBMC and 75.31614.7 for plasma vs. PBMC. SCR were re-analyzed after subjects were stratified according to VL, ART status, CD4 counts, and duration of HIV infection. When compared to subjects with a VL of .5,000 copies/ml, subjects with a lower VL had a greater sequence discordance between plasma and either DBS or PBMC. The mean SCR for DBS vs. plasma was 72.0 when the VL was ,5,000 copies/ml, and 88.8 when the VL was 5,000 copies/ml (p = 0.002). The plasma vs. PBMC SCR was 65.7 vs 80.6 (p = 0.042) in the same low/high viral load stratification. No significant difference in the SCR was found when comparing sequences from DBS and those from PBMC at different VL (75.3 vs 80.9 , p = 0.325) (Figure 1a). Current ART use also had a significant impact on the SCR. When stratified according to ART status, the mean SCRs in the ?off-ART group (ART-naive or prior ART) were 86.3 , 80.7 and 78.6 for the pairs of DBS vs. plasma, DBS vs. PBMC and plasma vs. PBMC respectively. In contrast, the corresponding SCRs in the on-ART groups were 67.2 , 70.5 and 59.9 respectively. The inter-group differences were statistically significant for plasma vs.DBS and plasma vs. PBMCs pairs (p = 0.007and 0.041 respectively). Although not statistically significant, the DBS vs. PBMC comparison showed a trend to greater differences in SCR when the patients were off-ART (p = 0.146) (Figure 1b). Although ART exposure appeared influences SCR, multivariate analysis demonstrated that VL level was the only independent correlate of SCR when comparing DBS with plasma. The duration of HIV infection was correlated with the SCRs found when the specimens were compared (Figure 1C). For example, a significantly higher SCR between DBS- and plasmabased HIV genotypes were observed in patients with an infection of #2 years as compared to those infected by HIV for longer periods (90.5 vs 82.8 , p = 0.044) . This finding was also seen when plasma and PBMC genotypes were compared. (SCR 84.0 vs 74.8 , p = 0.05). The same SCR trend was observed for the DBSvs. PBMC comparison although the difference was not significant (p = 0.08).Subjects and specimenAfter informed consent, 17 HIV-1 positive subjects provided an EDTA anti-coagulated blood specimen. A FACSCalibur flow cytometer (BD Biosciences, USA) was used for CD4+ cell enumeration. DBS were prepared by pipetting 75 ml/spot of whole blood onto Whatman 903H filter paper (Whatman Inc, Florham Park, USA). Each card was air-d.