Email this article 
Hide
Show all
Original Research
Validation of PHASE for deriving N-acetyltransferase 2 haplotypes in the Western Cape mixed ancestry population
Submitted: 28 January 2019 | Published: 17 December 2020
About the author(s)
Celeste Swart, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa; and, National Health Laboratory Service (NHLS), Groote Schuur Hospital, Cape Town, South AfricaSurita Meldau, National Health Laboratory Service (NHLS), Groote Schuur Hospital, Cape Town, South Africa; and, Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa
Chad M. Centner, Division of Medical Microbiology, University of Cape Town, Cape Town, South Africa; and, National Health Laboratory Service (NHLS), Medical Microbiology, Groote Schuur Hospital, Cape Town, South Africa
Adrian D. Marais, Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa
Fierdoz Omar, Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa
Abstract
Background: There is a shortage of data on the accuracy of statistical methods for the prediction of N-acetyltransferase 2 (NAT2) haplotypes in the mixed ancestry population of the Western Cape.
Objective: This study aimed to identify the NAT2 haplotypes and assess the accuracy of PHASE version 2.1.1 in assigning NAT2 haplotypes to a mixed ancestry population from the Western Cape.
Methods: This study was conducted between 2013 and 2016. The NAT2 gene was amplified and sequenced from the DNA of 100 self-identified mixed ancestry participants. Haplotyping was performed by molecular and computational techniques. Agreement was assessed between the two techniques.
Results: Haplotypes were assigned to 93 samples, of which 67 (72%) were ambiguous. Haplotype prediction by PHASE demonstrated 94.6% agreement (kappa 0.94, p < 0.001) with those assigned using molecular techniques. Five haplotype combinations (from 10 chromosomes) were incorrectly predicted, four of which were flagged as uncertain by the PHASE software. Only one resulted in the assignment of an incorrect acetylation phenotype (intermediate to slow), although the software flagged this for further analysis. The most common haplotypes were NAT2*4 (28%) followed by NAT2*5B (27.4%), NAT2*6A (21.5%) and NAT2*12A (7.5%). Four rare single nucleotide variants (c.589C>T, c.622T>C, c.809T>C and c.387C>T) were detected.
Conclusion: PHASE accurately predicted the phenotype in 92 of 93 samples (99%) from genotypic data in our mixed ancestry sample population, and is therefore a suitable alternative to molecular methods to individualise isoniazid therapy in this high burden tuberculosis setting.
Keywords
Metrics
Total abstract views: 2473Total article views: 2916