- Open Access
Comparative determination of HIV-1 co-receptor tropism by Enhanced Sensitivity Trofile, gp120 V3-loop RNA and DNA genotyping
© Prosperi et al; licensee BioMed Central Ltd. 2010
- Received: 9 March 2010
- Accepted: 30 June 2010
- Published: 30 June 2010
Trofile® is the prospectively validated HIV-1 tropism assay. Its use is limited by high costs, long turn-around time, and inability to test patients with very low or undetectable viremia. We aimed at assessing the efficiency of population genotypic assays based on gp120 V3-loop sequencing for the determination of tropism in plasma viral RNA and in whole-blood viral DNA. Contemporary and follow-up plasma and whole-blood samples from patients undergoing tropism testing via the enhanced sensitivity Trofile® (ESTA) were collected. Clinical and clonal geno2pheno[coreceptor] (G2P) models at 10% and at optimised 5.7% false positive rate cutoff were evaluated using viral DNA and RNA samples, compared against each other and ESTA, using Cohen's kappa, phylogenetic analysis, and area under the receiver operating characteristic (AUROC).
Both clinical and clonal G2P (with different false positive rates) showed good performances in predicting the ESTA outcome (for V3 RNA-based clinical G2P at 10% false positive rate AUROC = 0.83, sensitivity = 90%, specificity = 75%). The rate of agreement between DNA- and RNA-based clinical G2P was fair (kappa = 0.74, p < 0.0001), and DNA-based clinical G2P accurately predicted the plasma ESTA (AUROC = 0.86). Significant differences in the viral populations were detected when comparing inter/intra patient diversity of viral DNA with RNA sequences.
Plasma HIV RNA or whole-blood HIV DNA V3-loop sequencing interpreted with clinical G2P is cheap and can be a good surrogate for ESTA. Although there may be differences among viral RNA and DNA populations in the same host, DNA-based G2P may be used as an indication of viral tropism in patients with undetectable plasma viremia.
- CCR5 Antagonist
- Viral Tropism
- Position Specific Score Matrice
- Tropism Prediction
Maraviroc (MVC) is the first CCR5 antagonist approved for the treatment of HIV-1 infection  following the demonstration of its virological efficacy in treatment-experienced patients [2, 3]. There is reasonable expectation that MVC or other CCR5-antagonists can be even better administered to treatment-naïve patients due to a higher prevalence of CCR5-tropic (R5) HIV-1 in this population as compared to more advanced patients . Due to the lack of virologic activity against CXCR4-tropic (X4) virus , the administration of MVC is subject to the verification of an R5 virus population in the candidate patient. The enhanced sensitivity Trofile ® assay (ESTA) is the current gold standard phenotypic method for the determination of co-receptor tropism for the replicating viral population (i.e. plasma RNA), although other in-house or commercial tests are available, some of which use peripheral blood mononuclear cell (PBMC DNA) [6, 7]. The drawbacks of any phenotypic test include high costs, long turn-around time, and reduced efficiency in patients with low viremia. For this reason, there is a demand for a fast and cheap HIV-1 tropism assay to fully exploit CCR5 antagonists as a treatment option in clinical routine [8, 9].
Given that most of the determinants of viral co-receptor tropism are based on polymorphisms of the third hypervariable region (V3) of the gp120, an alternative to the phenotypic approach is the usage of machine learning tools based on viral genotypic information. So called in-silico or virtual phenotype models may be indeed convenient for clinical practice due to the reduction of costs and turn-around time. During the recent years, several prediction models have been studied, from the first simple rule based on the polymorphisms at V3 codons 11 and 25, to the position specific scoring matrices (PSSM), neural networks, support vector machines, random forests and logistic models [10–20]. Some of these studies identified additional factors possibly impacting viral tropism, such as viral subtype and CD4 cell counts [14, 16, 20]. Comparisons among genotypic and phenotypic tests have been carried out. The genotypic geno2pheno[coreceptor] system  was compared with the first generation Trofile ® and the TRT phenotypic assays , showing 86.5% and 79.7% concordance, respectively. A study comparing the predictive performance of geno2pheno[coreceptor], PSSM  and other methods against the first-generation Trofile ® assay, concluded that current default implementations of co-receptor prediction algorithms were inadequate for predicting CXCR4 co-receptor usage in clinical samples, due to inability to detect low-level X4 virus . Another study found the concordance among genotype-based predictors and first-generation Trofile ® being as high as 91% . Variable performance of in-silico models was shown when considering non-B HIV-1 variants [24–26].
Concerning the clinical validation of phenotypic assays, another recent work focused on the performance of the Trofile ® test in predicting the virological response to a short-term maraviroc exposure in HIV-infected patients . Concomitantly, a few attempts to unveil mutational patterns associated to selection by CCR5 antagonists or resistance have been carried out [28, 29], but the frequency and rate at which maraviroc resistance mutations emerge are not yet known.
The most awaited information is how in-silico tropism prediction models predict virological response to CCR5-antagonists, particularly when genotypic and phenotypic results disagree. In fact, although the ESTA should detect lower amounts of X4 virus compared to bulk genotyping, the X4 level threshold compromising in vivo CCR5-antagonist activity in vivo is currently unknown. In addition, ESTA cannot be performed at low or undetectable viral load (VL), while HIV-1 DNA genotyping can be easily performed in such cases; and this information, if adequately validated, might be easily employed for guiding treatment switches to CCR5 antagonists in virologically suppressed patients, due to toxicity or simplification issues. Finally, genotyping can also be used to detect HIV-1 mutations selected by MVC and possibly decreasing its effectiveness. Our study aimed at evaluating the accuracy of HIV-1 co-receptor tropism prediction by viral RNA and DNA genotyping, as well as the selection of V3 mutants in MVC-failing individuals.
Contemporary plasma and whole-blood samples were prospectively collected from HIV-infected patients followed up at a single centre of the Infectious Diseases Clinic of Catholic University of Sacred Heart in Rome, Italy, all failing antiretroviral treatment and potentially candidates for treatment with a CCR5-antagonists, in the period between November 2007 and July 2009 (n = 55). Some of these patients underwent tropism testing via the ESTA (n = 51). Follow-up plasma and whole-blood samples from these patients were also collected, regardless of MVC treatment.
Viral amplification and sequencing
Plasma RNA and whole-blood DNA were obtained from citrated blood by spin column extraction (Qiagen, Hilden, Germany). Plasma underwent 1-hour centrifugation at 23,000 x g at 4°C to concentrate virus prior to extraction. Whole blood was used without pre-processing steps. A 419-bp region encompassing the HIV-1 gp120 V3 domain was amplified by (RT)-PCR and sequenced in both strands by infrared-labelled primers on a Licor IR2 system. Plasma RNA was reverse transcribed with primer P151 (5'-CTACTTTATATTTATATAATTCAYTTCTC-3', coordinates 7661-7689 in the reference HXB2 genome). The reverse transcription reaction was run for 30 minutes at 37°C and included 10 μL of RNA extract, 50 mM Tris-HCl (pH 8.3 at 25°C), 75 mM KCl, 10 mM DTT, 3 mM MgCl2, 200μM each dNTP, 200 U ImProm II RT (Promega), 20 U RNasin (Promega) and 5 pmol primer P151. The cDNA obtained (one third of the reverse transcription mixture) or blood DNA (1 μg) extracted from whole-blood was amplified by a nested PCR protocol using primer P150 (5'-AATGTCAGCACAGTACAATGYACACAT-3', 6945-6971) and P151 in the outer amplification step and primer LR33 (5'-CAGTACAATGTACACATGGAAT-3', 6955-6976) and LR34 (5'-GAAAAATTCCCCTCCACAATT-3', 7353-7373) in the inner amplification step. Both outer and inner PCR mixtures contained 50 mM Tris-HCl (pH 9.0 at 25°C), 50 mM KCl, 2 mM MgCl2, 200 μM each dNTP, 1.25 U GoTaq polymerase (Promega) and 8 pmol each primer. The cycling profile was 20 seconds at 52°C, 40 seconds at 72°C and 30 seconds at 94°C for both steps but the number of cycles was 25 in the outer PCR and 32 in the inner PCR. The final product was sequenced by the IRD700-labelled sense primer IR25 (5'-GCTGTTAAATGGCAGTCTAGCAGAA-3', 7011-7035) and the IRD800-labelled antisense primer IR77 (5'-GAAAAATTCTCCTCYACAATTRA-3', 7351-7373) using the DYEnamic Direct Cycle Sequencing kit with 7-deaza-dGTP (GE Healthcare). Sequence contigs were assembled by the DNASTAR SeqMan II version 5.07 module. Only one PCR product per sample was subjected to standard population sequencing, expected to allow detection of minority species contributing at least 20% of the whole virus quasispecies.
Sequence analysis: tropism prediction and its evaluation
The geno2pheno[coreceptor] in-silico genotypic tropism prediction system was employed, using both clinical (requiring VL, nadir CD4 and CD8 count) and clonal interpretations at 10% false positive rate (FPR) , and by considering a clonal interpretation at the optimised cutoff of 5.75% FPR, based on analysis of clinical data from MOTIVATE and MERIT studies [30, 31]. Concordances among geno2pheno predictors and ESTA were assessed by Cohen's kappa statistic . The predictive ability of the systems was evaluated by receiver operating characteristic (ROC) analysis [33, 34].
The distance between nucleotide V3 sequences was calculated by the maximum composite likelihood , over a multiple alignment obtained via MUSCLE . Phylogenetic analysis was performed using the MEGA 4 software , estimating tree and branch support with a bootstrapped neighbour-joining method.
Analysis of selection of env mutations by MVC exposure was performed by considering patients with a RNA sequence before (the closest to the start date) and after (considering the latest after at least one month of therapy) MVC initiation. Amino acid mutations were extracted by local pairwise alignments against the HXB2 reference sequence. Fisher's exact test on frequency counts of individual mutations and pre- post-MVC strata was executed; obtained p-values were also corrected with Benjamini-Hochberg procedure.
The R mathematical programming suite was used to perform all statistical analyses .
Study population and samples
We processed 178 samples (99 plasma RNA and 79 whole-blood DNA) and successfully amplified 155 samples [78 (78.8%) plasma RNA and 77 (97.5%) whole-blood DNA] from 55 patients. By stratifying for contemporary VL, the rate of successful sequencing from RNA was 40%, 88.2%, and 96.2% at < = 50 cp/ml, 51-500 cp/ml, and >500 cp/ml respectively. The rate of successful sequencing from DNA was 95.1%, 100%, and 100% at < = 50 cp/ml, 50-500 cp/ml, and >500 cp/ml, respectively.
Patients' characteristics at the time of ESTA
No. of patients with available ESTA result
ESTA results, n (%)
Female gender, n (%)
Mean age (SD)
Median time from HIV-1 diagnosis, years (IQR)
Italian nationality, n (%)
Risk Factor, n (%)
CDC stage, n (%)
History of mono-dual NRTI therapy, n (%)
Median duration of ART exposure, years (IQR)
No. of drug switches (any change or interruption) experienced (IQR)
Patients previously exposed to drug class, n (%)
CD4 count, median cells/mm3 (IQR)
HIV-1 RNA load, median log10 cp/ml (IQR)
CD4 count at nadir, median cells/mm3 (IQR)
HIV-1 RNA load at zenith, median Log10 cp/ml (IQR)
In a cross sectional crude analysis, using dichotomized ESTA results (R5-tropic versus X4- plus dual/mixed-tropic), we did not find any significant association with contemporary patients characteristics, except for the time from HIV-1 diagnosis [median (IQR) 16 years (13-18) for R5, and 18 years (17-22) for X4 or dual/mixed-tropic isolates p = 0.007], and nadir CD4 count [median (IQR) 138 cells/mm3 (31-200) for R5 and 14 cells/mm3 (6-41) for X4 or dual/mixed-tropic isolates p = 0.05].
Comparison between geno2pheno[coreceptor]classification of plasma HIV-1 RNA, DNA V3 sequences and ESTA tropism results
Performance evaluation of ESTA X4-dual/mixed tropism prediction using geno2pheno clinical and clonal at 10% FPR, or the clonal at the optimised 5.75% FPR interpretation of contemporary viral gp120 V3 DNA or RNA genotyping
geno2pheno interpretation mode (FPR)
plasma RNA (n = 35)
Clonal optim. (5.75% )
whole-blood DNA (n = 17)
Clonal optim. (5.75% )
Comparison of HIV-1 V3 RNA and DNA sequence population and their inferred coreceptor tropisms
The median intrapatient distance among HIV RNA or DNA sequences was, as expected, significantly smaller than the interpatient distance (see Additional files 1 and 2). However, the intrapatient variability with RNA (0.007) or with DNA (0.015) sequences was significantly lower than the intrapatient variability between RNA and DNA sequences (0.023), suggesting that the HIV-1 V3 DNA and RNA populations may contribute different information and complement each other in a patient. Phylogenetic analysis revealed that clusters of sequences with high support values (bootstrap >90%) corresponded to sequences drawn from the same patients, and there were no clusters composed of sequences from different patients. Clusters were preferentially (but not exclusively) composed of either paired RNA or DNA samples (phylogenetic tree is available as Additional file 1).
We next investigated whether a certain degree of intrapatient diversity between DNA and RNA populations results in a different co-receptor tropism prediction. When comparing geno2pheno clinical interpretation based on paired HIV-1 V3 DNA-RNA sequences obtained from 29 distinct patients, we observed 35/40 (87.5%) concordant predictions, 3/40 (7.5%) false positives, and 2/40 (5%) false negatives, using RNA-predicted X4-tropism as the reference outcome. The kappa statistics yielded a strength of agreement of 0.74 (95% CI: 0.53-0.95), with a Fisher's p-value < 0.0001. The AUROC obtained by predicting the geno2pheno RNA tropism from contemporary DNA sequences was 0.875.
Viro-immunological follow up
Out of 28 patients with an R5-tropic virus by ESTA subsequently starting MVC, 22 had an available virological follow up at 12 weeks. All patients had a VL below 50 cp/ml, except two patients with 120 and 292 cp/ml. Virological follow-up at 24 weeks was available for 19 patients. All of these had a VL below 50 cp/ml, except two patients (different from those at three months) with a VL of 2,003 and 88 cp/ml.
Immunological follow-up at 12 weeks was available for 20 out of the 28 patients that started MVC. The median (IQR) CD4 increase was 84 (range -9 to 165) cells/mm3. A Wilcoxon test showed that this increase from baseline was statistically significant (p = 0.019). After 24 weeks of therapy, immunological follow-up was available for 19 patients. The median (IQR) CD4 increase was 46 (-9 to 143) cells/mm3(p = 0.053).
Evolution of V3 sequences during MVC therapy
In the present study, viral tropism prediction by the HIV-1 gp120 V3-loop RNA genotype-based geno2pheno[coreceptor] interpretation in treatment-experienced subjects proved to be a valid alternative to ESTA tropism, yielding an AUROC of 0.83 using the clinical interpretation. The clonal interpretation showed a lower AUROC, although the difference with the clinical interpretation showed only a trend, probably due to the limited sample size. The geno2pheno[coreceptor] website suggests that only the clonal interpretation should be used for treatment-experienced patients, since the clinical system was trained only on treatment-naïve patients. However, our results show that the clinical interpretation is better (at least not inferior) than the clonal interpretation even in treatment-experienced patients. It is important to highlight the fact that the clinical interpretation needs contemporary VL, nadir CD4 and CD8 information in order to work properly. Interestingly, the clinical mode improved sensitivity much more than specificity with respect to the clonal mode. The sensitivity of V3 RNA genotyping followed by clonal geno2pheno[coreceptor] interpretation has been recently shown to increase by testing multiple aliquots of the plasma RNA extract , likely as a consequence of stochastic fluctuations of a minority X4 populations. Although repeated testing has not yet been established as a standard procedure, it is expected that its implementation would result in a higher concordance between tropism results obtained by V3 genotyping and the reference ESTA.
Another area under investigation is the optimal FPR to use with geno2pheno[coreceptor] as a cut-off for classification of R5 and X4-D/M viruses. Recent retrospective analysis of the MERIT trial data indicated that the most accurate FPR cutoff may be in the range from 2% to 5.75%, when triplicate PCR testing and fully automated sequencing is used . However, there is currently no consensus on which cut-off to use in clinical practice. As for any classification test aimed at orienting a clinical intervention, the trade-off between specificity and sensitivity must be taken into account. A higher cut-off such as the 10% FPR was originally proposed  and is still recommended by the expert panel developing and maintaining the geno2pheno[coreceptor] system (see the German recommendations ). Using a higher cut-off translates into a conservative attitude, i.e. a lower probability to treat with maraviroc patients who may not benefit from it, at the expense of a higher probability not to treat patients who may benefit from the drug. It remains to be established whether this lower cut-off can be clinically more convenient in the genotypic screening of the general HIV patient population candidate to treatment with maraviroc.
One major advantage with genotyping is that even patients with non-reportable ESTA can be given prediction of tropism. By definition, ESTA is subject to a larger proportion of failures with respect to genotype due to virus polymorphisms invalidating the cloning procedure and an inherently lower rate of reverse transcription of a far larger virus genome region. Moreover, our efficiency of RNA genotyping at VL between 50 and 500 cp/ml (where ESTA is not even attempted) was 88%, whilst that of DNA genotyping was 100% at these viral loads.
Interestingly, in-silico tropism prediction using whole-blood DNA genotyping may be a solution when considering treatment switch to a CCR5-antagonists for patients with undetectable viral load. The perspective for treatment switches in these patients may be attractive, because of the good tolerability of MVC and because patients that are not at a late stage of disease are more likely to harbour a CCR5- rather than a CXCR4-tropic virus . In these cases ESTA cannot be used, and RNA genotyping is often not sufficiently efficient. The performance of DNA-based clinical geno2pheno in predicting the ESTA result was comparable to that of RNA-based clinical geno2pheno (AUROC = 0.86). Thus, the use of V3 DNA sequence data for predicting co-receptor tropism definitely warrants further investigation as an appealing alternative to RNA.
As expected, we found some differences when comparing paired DNA and RNA sequences, consistent with the notion that the archived population may not correspond to the most prevalent virus in plasma, whose source are the productively infected cells. However, when comparing DNA and RNA tropism prediction by looking at contemporary samples, the degree of agreement was good, implying that such minor differences may not commonly translate into inappropriate indications. It remains to be established whether an X4 virus population detected in PBMC DNA in the context of R5 virus in plasma RNA can impair response to maraviroc. Although reported on a limited number of cases, this did not appear to be the case in the French GenoTropism study [41, 42].
HIV-1 tropism determination via plasma viral V3 RNA genotyping coupled with geno2pheno interpretation may represent a valid alternative to ESTA. The clinical validation of genotypic determination of viral tropism has been recently performed using retrospective samples from the MOTIVATE study  as well as in the GenoTropism study where the genotypic tropism test was able to predict response to maraviroc even in the group of patients with an R5 virus population as detected by standard Trofile® [41, 42]. As shown here, RNA tropism genotyping carries the advantage of a higher efficiency of tropism determination in patients with low copy number detectable viral loads. In addition to that, in perspective, DNA-based tropism prediction could be used in patients with undetectable VL who are candidates for treatment simplification/switch to a CCR5-antagonist. This option can support a more effective use of this class of agents at earlier stages when the probability of harbouring an R5 virus population is maximal. However, further investigations to unveil the evolutionary relationships between DNA and RNA populations are advisable before DNA genotyping can be indicated in clinical practice. In this context, ultra-deep sequencing studies may be appropriate to dissect the dynamics and role of DNA and RNA minority variants [43, 44]. Most importantly, clinical validation of the use of HIV-1 genotyping, particularly with proviral DNA, for tropism assignment is also required before its widespread implementation.
This work has been partly supported by EU-funded projects DynaNets (grant #233847) and CHAIN (grant #223131).
- Sayana S, Khanlou H: Maraviroc: a new CCR5 antagonist. Expert Rev Anti Infect Ther. 2009, 7 (1): 9-19. 10.1586/14787220.127.116.11.View ArticlePubMedGoogle Scholar
- Heera J, Ive P, Botes M, Dejesus E, Mayer H, Goodrich J, Clumeck N, Cooper DA, Walmsley S, Craig C, Reeves J, van der Ryst E, Saag M: The MERIT study of maraviroc in antiretroviral-naive patients with R5 HIV-1: 96-week results [abstract]. 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention. Cape Town, South Africa. 2009Google Scholar
- Fätkenheuer G, Nelson M, Lazzarin A, Konourina I, Hoepelman AI, Lampiris H, Hirschel B, Tebas P, Raffi F, Trottier B, Bellos N, Saag M, Cooper DA, Westby M, Tawadrous M, Sullivan JF, Ridgway C, Dunne MW, Felstead S, Mayer H, van der Ryst E, MOTIVATE 1 and MOTIVATE 2 Study Teams: Subgroup analyses of maraviroc in previously treated R5 HIV-1 infection. N Engl J Med. 2008, 359 (14): 1442-55. 10.1056/NEJMoa0803154.View ArticlePubMedGoogle Scholar
- Vandekerckhove L, Verhofstede C, Vogelaers D: Maraviroc: perspectives for use in antiretroviral-naive HIV-1-infected patients. J Antimicrob Chemother. 2009, 63 (6): 1087-96. 10.1093/jac/dkp113.View ArticlePubMedGoogle Scholar
- Saag M, Goodrich J, Fätkenheuer G, Clotet B, Clumeck N, Sullivan J, Westby M, van der Ryst E, Mayer H, A4001029 Study Group: A double-blind, placebo-controlled trial of maraviroc in treatment-experienced patients infected with non-R5 HIV-1. J Infect Dis. 2009, 199 (11): 1638-47. 10.1086/598965.View ArticlePubMedGoogle Scholar
- Braun P, Wiesmann F: Phenotypic assays for the determination of coreceptor tropism in HIV-1 infected individuals. Eur J Med Res. 2007, 12 (9): 463-72.PubMedGoogle Scholar
- Coakley E, Reeves JD, Huang W, Mangas-Ruiz M, Maurer I, Harskamp AM, Gupta S, Lie Y, Petropoulos CJ, Schuitemaker H, van't Wout AB: Comparison of human immunodeficiency virus type 1 tropism profiles in clinical samples by the Trofile and MT-2 assays. Antimicrob Agents Chemother. 2009, 53 (11): 4686-93. 10.1128/AAC.00229-09.PubMed CentralView ArticlePubMedGoogle Scholar
- Rose JD, Rhea AM, Weber J, Quiñones-Mateu ME: Current tests to evaluate HIV-1 coreceptor tropism. Curr Opin HIV AIDS. 2009, 4 (2): 136-42. 10.1097/COH.0b013e328322f973.View ArticlePubMedGoogle Scholar
- Soriano V, Perno CF, Kaiser R, Calvez V, Gatell JM, di Perri G, Pillay D, Rockstroh J, Geretti AM: When and how to use maraviroc in HIV-infected patients. AIDS. 2009, 23 (18): 2377-85. 10.1097/QAD.0b013e328332d32d.View ArticlePubMedGoogle Scholar
- De Jong JJ, De Ronde A, Keulen W, Tersmette M, Goudsmit J: Minimal requirements for the human immunodeficiency virus type 1 V3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution. J Virol. 1992, 66 (11): 6777-6780.PubMed CentralPubMedGoogle Scholar
- Resch W, Hoffman N, Swanstrom R: Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology. 2001, 288: 51-62. 10.1006/viro.2001.1087.View ArticlePubMedGoogle Scholar
- Jensen MA, van't Wout AB: Predicting HIV-1 coreceptor usage with sequence analysis. AIDS Rev. 2003, 5: 104-112.PubMedGoogle Scholar
- Pillai S, Good B, Richman D, Corbeil J: A new perspective on V3 phenotype prediction. AIDS Res Hum Retroviruses. 2003, 19: 145-149. 10.1089/088922203762688658.View ArticlePubMedGoogle Scholar
- Sander O, Sing T, Sommer I, Low AJ, Cheung PK, Harrigan PR, Lengauer T, Domingues FS: Structural descriptors of gp120 V3 loop for the prediction of HIV-1 coreceptor usage. PLoS Comput Biol. 2007, 3 (3): e58-10.1371/journal.pcbi.0030058.PubMed CentralView ArticlePubMedGoogle Scholar
- Xu S, Huang X, Xu H, Zhang C: Improved prediction of coreceptor usage and phenotype of HIV-1 based on combined features of V3 loop sequence using random forest. J Clin Microbiol. 2007, 45 (5): 441-6.Google Scholar
- Sing T, Low AJ, Beerenwinkel N, Sander O, Cheung PK, Domingues FS, Büch J, Däumer M, Kaiser R, Lengauer T, Harrigan PR: Predicting HIV coreceptor usage on the basis of genetic and clinical covariates. Antivir Ther. 2007, 12 (7): 1097-106.PubMedGoogle Scholar
- Lengauer T, Sander O, Sierra S, Thielen A, Kaiser R: Bioinformatics prediction of HIV coreceptor usage. Nat Biotechnol. 2007, 25 (12): 1407-10. 10.1038/nbt1371.View ArticlePubMedGoogle Scholar
- Lamers SL, Salemi M, McGrath MS, Fogel GB: Prediction of R5, X4, and R5X4 HIV-1 coreceptor usage with evolved neural networks. IEEE/ACM Trans Comput Biol Bioinform. 2008, 5 (2): 291-300. 10.1109/TCBB.2007.1074.PubMed CentralView ArticlePubMedGoogle Scholar
- Boisvert S, Marchand M, Laviolette F, Corbeil J: HIV-1 coreceptor usage prediction without multiple alignments: an application of string kernels. Retrovirology. 2008, 5: 110-10.1186/1742-4690-5-110.PubMed CentralView ArticlePubMedGoogle Scholar
- Prosperi MC, Fanti I, Ulivi G, Micarelli A, De Luca A, Zazzi M: Robust supervised and unsupervised statistical learning for HIV type 1 coreceptor usage analysis. AIDS Res Hum Retroviruses. 2009, 25 (3): 305-14. 10.1089/aid.2008.0039.View ArticlePubMedGoogle Scholar
- Skrabal K, Low AJ, Dong W, Sing T, Cheung PK, Mammano F, Harrigan PR: Determining human immunodeficiency virus coreceptor use in a clinical setting: degree of correlation between two phenotypic assays and a bioinformatic model. J Clin Microbiol. 2007, 45 (2): 279-84. 10.1128/JCM.01118-06.PubMed CentralView ArticlePubMedGoogle Scholar
- Low AJ, Dong W, Chan D, Sing T, Swanstrom R, Jensen M, Pillai S, Good B, Harrigan PR: Current V3 genotyping algorithms are inadequate for predicting X4 co-receptor usage in clinical isolates. AIDS. 2007, 21 (14): F17-24. 10.1097/QAD.0b013e3282ef81ea.View ArticlePubMedGoogle Scholar
- Raymond S, Delobel P, Mavigner M, Cazabat M, Souyris C, Sandres-Sauné K, Cuzin L, Marchou B, Massip P, Izopet J: Correlation between genotypic predictions based on V3 sequences and phenotypic determination of HIV-1 tropism. AIDS. 2008, 22 (14): F11-6. 10.1097/QAD.0b013e32830ebcd4.View ArticlePubMedGoogle Scholar
- Garrido C, Roulet V, Chueca N, Poveda E, Aguilera A, Skrabal K, Zahonero N, Carlos S, García F, Faudon JL, Soriano V, de Mendoza C: Evaluation of eight different bioinformatics tools to predict viral tropism in different human immunodeficiency virus type 1 subtypes. J Clin Microbiol. 2008, 46 (3): 887-91. 10.1128/JCM.01611-07.PubMed CentralView ArticlePubMedGoogle Scholar
- Raymond S, Delobel P, Mavigner M, Ferradini L, Cazabat M, Souyris C, Sandres-Sauné K, Pasquier C, Marchou B, Massip P, Izopet J: Prediction of HIV type 1 subtype C tropism by genotypic algorithms built from subtype B viruses. J Acquir Immune Defic Syndr. 2010, 53 (2): 167-75. 10.1097/QAI.0b013e3181c8413b.View ArticlePubMedGoogle Scholar
- Raymond S, Delobel P, Mavigner M, Cazabat M, Souyris C, Encinas S, Sandres-Sauné K, Pasquier C, Marchou B, Massip P, Izopet J: Genotypic prediction of human immunodeficiency virus type 1 CRF02-AG tropism. J Clin Microbiol. 2009, 47 (7): 2292-4. 10.1128/JCM.02439-08.PubMed CentralView ArticlePubMedGoogle Scholar
- Genebat M, Ruiz-Mateos E, León JA, González-Serna A, Pulido I, Rivas I, Ferrando-Martínez S, Sánchez B, Muñoz-Fernández MA, Leal M: Correlation between the Trofile test and virological response to a short-term maraviroc exposure in HIV-infected patients. J Antimicrob Chemother. 2009, 64 (4): 845-9. 10.1093/jac/dkp293.View ArticlePubMedGoogle Scholar
- Anastassopoulou CG, Ketas TJ, Klasse PJ, Moore JP: Resistance to CCR5 inhibitors caused by sequence changes in the fusion peptide of HIV-1 gp41. Proc Natl Acad Sci USA. 2009, 106 (13): 5318-23. 10.1073/pnas.0811713106.PubMed CentralView ArticlePubMedGoogle Scholar
- Moore JP, Kuritzkes DR: A pièce de resistance: how HIV-1 escapes small molecule CCR5 inhibitors. Curr Opin HIV AIDS. 2009, 4 (2): 118-24. 10.1097/COH.0b013e3283223d46.PubMed CentralView ArticlePubMedGoogle Scholar
- McGovern R, Dong W, Zhong X, Knapp D, Thielen A, Chapman D, Lewis M, James I, Valdez H, Harrigan R: Population-based Sequencing of the V3-loop Is Comparable to the Enhanced Sensitivity Trofile Assay in Predicting Virologic Response to Maraviroc of Treatment-naï¿½ve Patients in the MERIT Trial [abstract]. 17th Conference on Retroviruses and Opportunistic Infections. San Francisco, USA. 2010Google Scholar
- Harrigan PR, McGovern R, Dong W, Thielen A, Jensen M, Mo T, Chapman D, Lewis M, James I, Ellery S, Heera J, Valdez H: Screening for HIV Tropism using Population based V3 Genotypic Analysis: a Retrospective Virologic Outcome Analysis Using Stored Plasma Screening Samples from the MOTIVATE Studies of Maraviroc in Treatment Experienced Patients [abstract]. 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention. Cape Town, South Africa. 2009Google Scholar
- Fleiss JL: Measuring nominal scale agreement among many raters. Psychological Bulletin. 1971, 76 (5): 378-382. 10.1037/h0031619.View ArticleGoogle Scholar
- Lasko TA, Bhagwat JG, Zou KH, Ohno-Machado L: The use of receiver operating characteristic curves in biomedical informatics. Journal of Biomedical Informatics. 2005, 38 (5): 404-415. 10.1016/j.jbi.2005.02.008.View ArticlePubMedGoogle Scholar
- DeLong ER, DeLong DM, Clarke-Pearson DL: Comparing the Areas Under Two or More Correlated Receiver Operating Characteristics Curves: A Nonparametric Approach. Biometrics. 1988, 44: 837-845. 10.2307/2531595.View ArticlePubMedGoogle Scholar
- Tamura K, Nei M, Kumar S: Prospects for inferring very large phylogenies by using the neighbor-joining method. Proceedings of the National Academy of Sciences USA. 2004, 101: 11030-11035. 10.1073/pnas.0404206101.View ArticleGoogle Scholar
- Edgar RC: MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004, 32 (5): 1792-1797. 10.1093/nar/gkh340.PubMed CentralView ArticlePubMedGoogle Scholar
- Tamura K, Dudley J, Nei M, Kumar S: MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution. 2007, 24: 1596-1599. 10.1093/molbev/msm092.View ArticlePubMedGoogle Scholar
- R Development Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Wien, Österreich. 2008, ISBN 3-900051-07-0, [http://www.R-project.org]Google Scholar
- Harrigan PR: The influence of PCR amplification variation on the ability of population-based PCR to detect non-R5 HIV [abstract]. 8th European HIV Drug Resistance Workshop, Sorrento, Italy. 2010Google Scholar
- German recommendations for determining HIV-1 coreceptor usage. [http://www.viro.med.uni-erlangen.de/nrz/recommendation080324.pdf]
- Morand-Joubert L, Flandre P, Soulié C, Charpentier C, Desbois D, Ruffault A, Marcelin AG, Masquelier B, Calvez V, the ANRS AC11 Resistance Study Group: Evolution of coreceptor tropism and V3 loop resistance mutations in the blood cellular reservoir after introduction of maraviroc. XVIII HIV Drug Resistance Workshop, Fort Myers, Florida. 2009Google Scholar
- Recordon-Pinson P, Descamps D, Soulie C, Lazrek M, Charpentier C, Montes B, Trabaud M-A, Cottalorda J, Amiel C, Morand-Joubert L, Tamalet C, Desbois D, Mace M, Ferre V, Vabret A, Ruffault A, Fleury H, Izopet J, Brun-Vezinet F, Marcelin AG, Reynes J, Flandre P, Calvez V, Masquelier B, ANRS AC11 Resistance Study Group: Genotypic prediction of HIV-1 tropism: correlation with the virological response to maraviroc and genotypic evolution on maraviroc therapy in the GenoTropism study. 12th European AIDS Conference (EACS), Cologne, Germany. 2009Google Scholar
- Archer J, Braverman MS, Taillon BE, Desany B, James I, Harrigan PR, Lewis M, Robertson DL: Detection of low-frequency pretherapy chemokine (CXC motif) receptor 4 (CXCR4)-using HIV-1 with ultra-deep pyrosequencing. AIDS. 2009, 23 (10): 1209-18. 10.1097/QAD.0b013e32832b4399.PubMed CentralView ArticlePubMedGoogle Scholar
- Tsibris AM, Korber B, Arnaout R, Russ C, Lo CC, Leitner T, Gaschen B, Theiler J, Paredes R, Su Z, Hughes MD, Gulick RM, Greaves W, Coakley E, Flexner C, Nusbaum C, Kuritzkes DR: Quantitative deep sequencing reveals dynamic HIV-1 escape and large population shifts during CCR5 antagonist therapy in vivo. PLoS One. 2009, 4 (5): e5683-10.1371/journal.pone.0005683.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.