Lack of association between STK39 and hypertension in the Chinese population


ORIGINAL ARTICLE

Lack of association between STK39 and hypertension
in the Chinese population
J Xu1,2,6, L-D Ji1,3,6, L-N Zhang2, C-Z Dong2, L-J Fei2, S Hua1, J-Y Tsai4 and Y-P Zhang1,5

Genome-wide association study (GWAS) has identified serine/threonine kinase 39 (STK39) as a candidate gene for hypertension. A
replication study provided supporting evidence that STK39 functional polymorphism rs35929607 was associated with hypertension.
Recently, another study also showed rs6749447 within the STK39 was associated with blood pressure responses. However, these
studies were all conducted in Caucasians. Thus, we carried out a case–control study to test whether STK39 is a common candidate
gene for hypertension, and to examine the interaction of genetic factors and non-genetic risk factors in the Chinese population.
Thousand twenty four hypertensive cases and 1024 controls were genotyped for five polymorphisms. Four single-nucleotide
polymorphisms (SNPs) are located within STK39, and rs4977950, the SNP that showed the strongest signal is located in a gene
desert. Results indicated that none of these SNPs was associated with hypertension in the Chinese population. Logistic regression
analysis found body mass index (BMI) and triglyceride level were higher in the hypertension group when compared with the
control group. Multifactor dimensionality reduction analysis indicated that the interaction between BMI and rs4977950 may have
an impact on hypertension. Taken together, the present study found no evidence that STK39 was associated with hypertension
in the Chinese population. Instead, non-genetic risk factors such as BMI have an important role in Chinese hypertensive subjects,
and the ‘missing inheritability’ requires further investigation.

Journal of Human Hypertension (2013) 27, 294–297; doi:10.1038/jhh.2012.46; published online 15 November 2012

Keywords: genome-wide association study; serine/threonine kinase 39; polymorphism; interaction

INTRODUCTION
Genome-wide association study (GWAS) has identified serine/
threonine kinase 39 (STK39) as a candidate gene for hypertension
in the American Old Order Amish.1 Recently, a replication study in
two Sweden cohorts showed that the STK39 functional
polymorphism rs35929607A4G was associated with blood
pressure (BP) and hypertension.2 Consequently, STK39 becomes
a very important ‘susceptibility gene’ for hypertension.
STK39 encodes a serine/threonine kinase known as a Ste20-

related proline–alanine-rich kinase (SPAK). When phosphorylated
by the WNK1 (WNK lysine-deficient protein kinase 1) and WNK4
(WNK lysine-deficient protein kinase 4), SPAK can bind and
phosphorylate two Naþ-dependent cation chloride cotranspor-
ters, including furosemide-sensitive Naþ/Kþ/2Cl� channel
(NKCC2) and thiazide-sensitive Naþ/Cl� channel (NCC).3–5 This
pathway has a critical role in salt homeostasis in renal physiology
and is strongly implicated in BP regulation and hypertension
development.3 Recently, Donner and colleagues selected 19
single-nucleotide polymorphisms (SNPs) from five GWASs and
studied their association with BP responses to four different
antihypertensive drug monotherapies in Finnish population. The
result indicated only rs6749447 (within the STK39 gene) was
significantly associated with BP responses.6

However, these studies were all conducted in Caucasians, and
whether it is a common candidate gene for hypertension in other
populations is not clear. Therefore, we carried out a case–control

study and an interaction analysis to verify whether STK39 is
associated with hypertension in the Chinese population.

METHODS
Study population
The participants were chosen from our established community-based
epidemiology study of common diseases. With informed consent, we
collected 410 000 health records. Subsequently, participants who fulfilled
the following criteria were put into our database: 30–75 years old, Han
Chinese, living in Ningbo City (East side of China) for at least three
generations and do not have a migration history. Thousand twenty four
essential hypertension cases and 1024 controls were chosen from this
database. The hypertensive subjects and control subjects were matched
for age and gender. In addition, only subjects who do not have
cardiovascular disease or other major chronic illnesses according to their
health records were included in the control group.

BP and clinical parameters
BP measurements were conducted in the morning. After the participant
had been in the sitting position for 10 min, three BP measurements were
obtained at 5-min intervals using standard mercury sphygmomanometer,
and the average of last two measurements was taken as the BP for that
participant. Hypertension in this study is defined as a sitting systolic BP
(SBP) X140mm Hg and/or a diastolic BP (DBP) X90 mmHg, or self-reported
use of antihypertension medication. Patients with secondary hypertension
were excluded. Normal BP is defined with SBP p120 mm Hg and DBP
p80 mm Hg.

1State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology, Chinese Academy of Sciences, Kunming, China; 2Department of Public Health, School of
Medicine, Ningbo University, Ningbo, China; 3Institute of Biochemistry, School of Medicine, Ningbo University, Ningbo, China; 4Center for Cardiovascular Genetics, Brown
Foundation Institute of Molecular Medicine, University of Texas Health Science Center at Houston, Houston, TX, USA and 5Laboratory for Conservation and Utilization of Bio-
resource, Yunnan University, Kunming, China. Correspondence: Professor Y-P Zhang, State Key Laboratory of Genetic resources and Evolution, Kunming Institute of Zoology,
Chinese Academy of Sciences, No.32 Jiaochang Donglu, Kunming 650223, China.
E-mail: zhangyp@mail.kiz.ac.cn
6These authors contributed equally to this work.
Received 3 September 2012; accepted 28 September 2012; published online 15 November 2012

Journal of Human Hypertension (2013) 27, 294–297
& 2013 Macmillan Publishers Limited All rights reserved 0950-9240/13

www.nature.com/jhh

http://dx.doi.org/10.1038/jhh.2012.46
mailto:zhangyp@mail.kiz.ac.cn
http://www.nature.com/jhh


Blood samples were collected with informed consent. Subsequently,
total cholesterol (TC), high-density lipoprotein (HDL), and triglyceride (TG)
were measured from these blood samples. Simultaneously, clinical infor-
mation including body mass index (BMI), smoking habit and alcohol
abuse were also obtained. The protocol of this study was reviewed and
approved by the ethics committees of Kunming Institute of Zoology,
Chinese Academy of Sciences and Ningbo University.

SNP genotyping
Genomic DNA was obtained from the whole blood by standard phenol/
chloroform extraction. Genotyping was performed by the GenomeLab
SNPstream Genotyping System (Beckman Coulter, Fullerton, CA, USA)
according to the protocols provided by the manufacturer.7 Of the five
candidate SNPs, rs12692877 was replaced by a nearby SNP (rs4667569,
D0 ¼ 1, r2 ¼ 1 in CHB) due to unavailability of appropriate primers
(www.autoprimer.com). The primers for SNPstream were listed in
Supplementary Table 1. rs4977950 was genotyped by PCR-restriction
fragment length polymorphism using mismatched primers (forward primer
5’-CTGGCATTTGTATTACTTTC-3’, reverse primer 5’-TGAACAGGTCCCA
TACTC-3’), and the PCR products were digested by the restriction enzyme
SpeI (New England Biolabs, Beverly, MA, USA).

Statistics
Continuous variables are presented as the mean±s.d. and analyzed by
t-test between the two groups. Statistical analysis of allele and genotype
frequencies between case and control groups and among different sex
groups were performed by w2-test. Effect of confounding variables were
identified by logistic regression (SPSS 16.0, SPSS Inc., Chicago, IL, USA).
Hardy–Weinberg equilibrium was determined by the software PEDSTATS
V0.6.8 (http://www.sph.umich.edu/csg/abecasis/).8 Multifactor dimensionality
reduction (MDR) was used to identify and characterize interactions among the
SNPs and the non-genetic factors, including BMI, serum HDL, TC and TG level,
as well as frequency of smoking and alcohol abuse.9 The software used for
MDR is distributed in a JAVA platform with a graphical user interface and is
freely available (http://www.epistasis.org/mdr.html).
All tests were two-sided and P-values o0.05 were considered statistically

significant. For w2-test, the P-values were adjusted for the total number of
tested SNPs using the Bonferroni correction method (a ¼ 0.05/5¼ 0.01).

RESULTS
The baseline characteristics of the participants are summarized in
Table 1. In this study, 1024 hypertensive participants and 1024
controls, who have all been long-term residents of Ningbo City,
were analyzed. The male to female ratio was equal in both groups,
and mean age of hypertensive participants and controls were
similar, demonstrating that the experimental and control groups
were well-matched and are appropriate for the following analyses.
BMI, serum TG and TC level in the hypertensive group were

significantly greater than the control group (Po0.01). Furthermore,
significantly more subjects drank alcohol regularly in hypertensive

group than those in control group (Po0.01). However, HDL level
and smoking habit were similar in both groups.
As shown in Supplementary Figures S1–S5, the results of

SNPstream and PCR-restriction fragment length polymorphism
were clear and reliable. Table 2 shows the genotypes and minor
allele frequency of each SNP. The genotyping success rate was
99.7%, and all five SNPs did not deviate from Hardy–Weinberg
equilibrium (P40.05). In addition, with the prevalence, odds ratio,
and minor allele frequency in this study, the Genetic Power
Calculator (available online http://pngu.mgh.harvard.edu/Bpur-
cell/gpc/) indicated that the sample size is big enough to do case–
control analysis with 80% power.10 According to the w2-test
P-values and odds ratios, none of the SNPs was shown to associate
with hypertension, even after stratification by gender (Table 2).
Considering the effect of confounding variables, we further

carried out logistic regression analysis with genetic and non-
genetic factors. The result indicated that only BMI and TG were
associated with hypertension (Po0.001), and none of the SNPs
was positive (Table 3).
Finally, MDR was used to analyze the interaction among

different SNPs and non-genetic risk factors for hypertension. After
input the five SNPs together with information of TG, TC, HDL, BMI,
smoking and alcohol drinking, the software output the best model
for BMI with 10/10 cross-validation consistency, and ‘BMI,
rs4977950’ with 9/10 cross-validation consistency (Table 4).

DISCUSSION
STK39 encodes SPAK, and the WNK-SPAK-NKCC2/NCC pathway is
involved in salt renal reabsorption and BP homeostasis.3

Therefore, it is not surprising to see that SNPs within STK39
were identified to associate with hypertension through GWAS. To
avoid the possibility of false-positive results, Fava et al. conducted
a replication study in two independent Sweden cohorts.2 Their
results indicated that the functional rs35929607A4G
polymorphism was associated with higher SBP and DBP values
in the Malmö Diet and Cancer (n ¼ 5634), but not in the Malmö
Preventive Project (n ¼ 17 894). After stratification for sex, the
results remained statistically significant only in women.2

Table 1. Baseline characters of the investigated participants

Variables Case Control P-value

Number 1024 1024 NA
Sex (male) (%) 43.6 43.6 NA
Age (years) 57.23±7.24 57.01±7.39 0.41
TG (mM) 2.02±1.69 1.64±1.12 o0.01
HDL (mM) 1.41±0.35 1.40±0.32 0.81
TC (mM) 5.34±1.01 5.18±0.93 o0.01
BMI (kg m� 2) 24.65±3.25 23.22±2.88 o0.01
Smoking (%) 18.90 16.37 0.12
Alcohol drinking (%) 22.86 19.23 o0.01

Abbreviations: BMI, body mass index; HDL, high-density lipoprotein;
TC, total cholesterol; TG, triglyceride.

Table 2. Association statistics for five SNPs

SNP Group MAF P-value OR 95% CI

rs6749447 Case 0.30 0.65 1.03 0.90–1.18
Control 0.31
Male 0.31 0.49 1.07 0.88–1.31
Female 0.30 0.99 1.00 0.84–1.19

rs3754777 Case 0.26 0.80 1.02 0.89–1.17
Control 0.25
Male 0.26 0.63 0.95 0.77–1.17
Female 0.25 0.44 1.08 0.89–1.30

rs35929607 Case 0.44 0.82 1.01 0.90–1.15
Control 0.44
Male 0.45 0.83 0.98 0.81–1.18
Female 0.43 0.62 1.04 0.88–1.23

rs4667569 Case 0.47 0.60 0.97 0.86–1.09
Control 0.46
Male 0.47 0.80 1.02 0.85–1.23
Female 0.46 0.36 0.93 0.79–1.09

rs4977950 Case 0.22 0.07 0.87 0.75–1.01
Control 0.20
Male 0.20 0.30 0.88 0.70–1.12
Female 0.22 0.14 0.86 0.71–1.05

Abbreviations: CI, confidence interval; MAF, minor allele frequency;
OR, odds ratio; SNP, single-nucleotide polymorphism.

STK39 and hypertension
J Xu et al

295

& 2013 Macmillan Publishers Limited Journal of Human Hypertension (2013) 294 – 297

www.autoprimer.com
http://www.sph.umich.edu/csg/abecasis/
http://www.epistasis.org/mdr.html
http://pngu.mgh.harvard.edu/~purcell/gpc/
http://pngu.mgh.harvard.edu/~purcell/gpc/


The population studied in the initial GWAS from Wang et al.1 is
American Old Order Amish, which is a closed founder population
emigrated from Switzerland in the early 1700s. The study
populations in articles of Fava et al. and Donner et al. are from
Swede and Finland. These populations are all Caucasians and
close in geographical location. When we checked the details of
the positive SNPs (rs6749447, rs3754777 and rs4977950) in Wang’s
GWAS, we found that the minor allele frequency differs greatly
between Europeans and Asians. To determine whether STK39 is a
common candidate gene for hypertension, we set to find out
whether it is associated with hypertension in the Chinese
population, as China has now become a nation with rapidly
increasing hypertension prevalence.
We first studied two SNPs (rs35929607 and rs12692877) within

STK39, which were putative functional SNPs based on a multispecies
sequence alignment.1 However, neither SNP was significantly asso-
ciated with hypertension in the Chinese population. As the result
was quite different from Fava’s replication study,2 we additionally
genotyped two tag SNPs (rs6749447 and rs3754777) within STK39.
These two SNPs belong to LD bin 1 and bin 2. Both SNPs were
significant with SBP in Amish (Po0.0001), marginally significant in
population from diabetes genetics initiative (P ¼ 0.02 for rs6749447
and 0.03 for rs3754777), but not significant in populations from
Framingham Heart Study from, GenNet and Hutterites.1 However,
in the present study, these two tag SNPs were still not associated
with hypertension. We further genotyped rs4977950 as it has the
minimum P-value in Wang’s GWAS. This SNP rs4977950 is not a
STK39 polymorphism, but a SNP that is located in a gene desert
900 kb away from known genes. In our study, the P-value for
rs4977950 is 0.074, and the odds ratio is 1.15 (95% confidence
interval: 0.99–1.33). Even after stratification with gender, none of the
five SNPs showed association with hypertension.
Hereto, five genome-wide significant SNPs genotyped in this

study did not show a positive association with hypertension in the
Chinese population. Similarly, Cunnington et al.11 also did not find
any association between the three SNPs (rs6749447, rs3754777,

rs35929607) within STK39 and SBP or DBP from a cohort of 1372
British Caucasians. Furthermore, in Fava’s study, after exclusion of
2398 individuals from the Malmö Preventive Project cohort, who
also participated in the Malmö Diet and Cancer study, G allele of
rs35929607 was no longer associated with hypertension.2 Thus, it
is still questionable whether STK39 can be accepted as a common
susceptibility gene for hypertension.
Pickering and colleagues have initially suggested that hypertension

and BP are complex traits.12 Before the era of GWAS, epidemiological
studies have found dozens of risk factors, such as smoking, alcohol
abuse, excessive salt intake, obesity, mental stress and others to
associate with high BP.13,14 Although no positive SNP was found
within this study, we found interesting interactions between genetic
factors and non-genetic risk factors. High BMI and serum TG level
were already believed to be risk factors for hypertension,15,16 and
were also confirmed by logistic regression in current study. The MDR
analysis further demonstrated that these important risk factors
interacted with the genetic factor. Thus, the present interaction
analysis gave a little more information than the single genetic study.
The participants in recent GWASs are mainly from well-organized
cohorts (Global BPgen and Cohorts for Heart and Aging Research in
Genomic Epidemiology), which means the epidemiology data are
readily available.17–19 With the development of statistic methods for
evaluation of gene–environment interaction, more missing inherit-
ability will be found.20,21

All in all, studies from Wang et al., Fava et al. and Donner et al.
have demonstrated that STK39 is associated with hypertension or
BP regulation in three different Caucasian populations,1,2,6 but this
result could not be replicated in the Chinese population. In
addition, in our interaction analysis, a significant interaction was
found between genetic and non-genetic factors, demonstrating
that GWAS alone is not a sufficient indicator for hypertension. To
decipher causal factors leading to the development and the
pathogenesis of hypertension, future work will require analysis of
both genetic (epigenetic) and non-genetic (environmental) factors.

What is known about the topic
� Genome-wide association study has identified STK39 as a candidate
gene for hypertension.

� A replication study in two Sweden cohorts provided supporting
evidence that STK39 functional polymorphism rs35929607 was
associated with hypertension. Another study conducted in Finnish
population also showed rs6749447 within the STK39 was associated
with blood pressure responses.

What this study adds
� None of the SNPs within the STK39 was associated with hypertension
in the Chinese population.

� Logistic regression showed that only body mass index (BMI) and TG
were associated with hypertension.

� MDR analysis indicated that the interaction among BMI and rs4977950
may have an impact on hypertension.

� The present study found no evidence that STK39 was associated with
hypertension in the Chinese population. Instead, interaction analysis
indicated that non-genetic risk factors such as BMI have an important
role for Chinese hypertensive subjects, and the ‘missing inheritability’
requires further investigation.

Table 3. Logistic regression for genetic and non-genetic factors

Variables B P-value Exp (B) 95% CI

Gender � 0.122 0.326 0.885 0.693–1.129
Age 0.010 0.132 1.010 0.997–1.024
rs6749447 0.074 0.576 1.076 0.832–1.393
rs3754777 0.087 0.556 1.091 0.815–1.461
rs35929607 0.036 0.767 1.037 0.817–1.315
rs4667569 0.140 0.347 1.150 0.859–1.540
rs4977950 0.130 0.120 1.139 0.967–1.341
BMI 0.146 0.000 1.157 1.118–1.198
TG 0.158 0.000 1.171 1.073–1.278
HDL 0.285 0.087 1.330 0.959–1.845
TC 0.039 0.484 1.040 0.932–1.161
Smoking 0.227 0.152 1.255 0.919–1.713
Alcohol 0.257 0.088 1.293 0.963–1.735
Constant � 5.405 0.000 0.004

Abbreviations: B, b-regression coefficient; CI, confidence interval.

Table 4. MDR analysis of gene-environment interaction

Best model Testing accuracy Testing sensitivity Testing odds ratio Testing w2 Cross-validation consistency

BMI 0.58 0.42 2.14 (95% CI: 1.14–4.02) 5.74 (P ¼ 0.0166) 10/10
BMI, rs4977950 0.59 0.61 2.08 (95% CI: 1.15–3.77) 5.99 (P ¼ 0.0144) 9/10
BMI, TC, rs4977950 0.61 0.54 2.45 (95% CI: 2.00–2.99) 77.63 (Po0.0001) 5/10

Abbreviations: BMI, body mass index; CI, confidence interval; MDR, multifactor dimensionality reduction; TC, total cholesterol.

STK39 and hypertension
J Xu et al

296

Journal of Human Hypertension (2013) 294 – 297 & 2013 Macmillan Publishers Limited



CONFLICT OF INTEREST
The authors declare no conflict of interest.

ACKNOWLEDGEMENTS
We acknowledge Dr Ali J Marian for helpful suggestions to this study. This study was
supported by Zhejiang Natural Science Foundation (Y2100857), Ningbo Natural
Science Foundation (2011A610037, 2012A610237) and China Postdoctoral Science
Foundation (20100481399).

REFERENCES
1 Wang Y, O’Connell JR, McArdle PF, Wade JB, Dorff SE, Shah SJ et al. Whole-

genome association study identifies STK39 as a hypertension susceptibility gene.
Proc Natl Acad Sci USA 2009; 106(1): 226–231.

2 Fava C, Danese E, Montagnana M, Sjogren M, Almgren P, Engstrom G et al. Serine/
threonine kinase 39 is a candidate gene for primary hypertension especially in
women: results from two cohort studies in Swedes. J Hypertens 2011; 29(3): 484–491.

3 Vitari AC, Deak M, Morrice NA, Alessi DR. The WNK1 and WNK4 protein kinases
that are mutated in Gordon’s hypertension syndrome phosphorylate and activate
SPAK and OSR1 protein kinases. Biochem J 2005; 391(Pt 1): 17–24.

4 Moriguchi T, Urushiyama S, Hisamoto N, Iemura S, Uchida S, Natsume T et al.
WNK1 regulates phosphorylation of cation-chloride-coupled cotransporters via
the STE20-related kinases, SPAK and OSR1. J Biol Chem 2005; 280(52):
42685–42693.

5 Delpire E, Gagnon KB. SPAK and OSR1: STE20 kinases involved in the regulation of
ion homoeostasis and volume control in mammalian cells. Biochem J 2008; 409(2):
321–331.

6 Donner KM, Hiltunen TP, Hannila-Handelberg T, Suonsyrja T, Kontula K. STK39
variation predicts the ambulatory blood pressure response to losartan in hyper-
tensive men. Hypertens Res 2012; 35(1): 107–114.

7 Bell PA, Chaturvedi S, Gelfand CA, Huang CY, Kochersperger M, Kopla R et al.
SNPstream UHT: ultra-high throughput SNP genotyping for pharmacogenomics
and drug discovery. Biotechniques 2002; 74(Suppl: 70–72): 76–77.

8 Wigginton JE, Cutler DJ, Abecasis GR. A note on exact tests of Hardy-Weinberg
equilibrium. Am J Hum Genet 2005; 76(5): 887–893.

9 Motsinger AA, Ritchie MD. The effect of reduction in cross-validation intervals on
the performance of multifactor dimensionality reduction. Genet Epidemiol 2006;
30(6): 546–555.

10 Purcell S, Cherny SS, Sham PC. Genetic power calculator: design of linkage and
association genetic mapping studies of complex traits. Bioinformatics 2003; 19(1):
149–150.

11 Cunnington MS, Kay C, Avery PJ, Mayosi BM, Koref MS, Keavney B. STK39 poly-
morphisms and blood pressure: an association study in British Caucasians and
assessment of cis-acting influences on gene expression. BMC Med Genet 2009;
10: 135.

12 Pickering GW, Keen H, Rose G, Smith A. The nature of essential hypertension.
Lancet 1959; 2(7110): 1027–1030.

13 Horan MJ, Lenfant C. Epidemiology of blood pressure and predictors of hyper-
tension. Hypertension 1990; 15(2 Suppl): I20–I24.

14 Gavras I, Manolis A, Gavras H. Genetic epidemiology of essential hypertension.
J Hum Hypertens 1999; 13(4): 225–229.

15 Filipovsky J, Simon J, Chrastek J, Rosolova H, Haman P, Petrikova V. Changes of
blood pressure and lipid pattern during a physical training course in hypertensive
subjects. Cardiology 1991; 78(1): 31–38.

16 Bhan V, Yan RT, Leiter LA, Fitchett DH, Langer A, Lonn E et al. Relation between
obesity and the attainment of optimal blood pressure and lipid targets in high
vascular risk outpatients. Am J Cardiol 2010; 106(9): 1270–1276.

17 Kato N, Takeuchi F, Tabara Y, Kelly TN, Go MJ, Sim X et al. Meta-analysis of
genome-wide association studies identifies common variants associated with
blood pressure variation in east Asians. Nat Genet 2011; 43(6): 531–538.

18 Levy D, Ehret GB, Rice K, Verwoert GC, Launer LJ, Dehghan A et al. Genome-wide
association study of blood pressure and hypertension. Nat Genet 2009; 41(6):
677–687.

19 Newton-Cheh C, Johnson T, Gateva V, Tobin MD, Bochud M, Coin L et al. Genome-
wide association study identifies eight loci associated with blood pressure.
Nat Genet 2009; 41(6): 666–676.

20 Mukherjee B, Ahn J, Gruber SB, Chatterjee N. Testing gene-environment inter-
action in large-scale case-control association studies: possible choices and com-
parisons. Am J Epidemiol 2012; 175(3): 177–190.

21 Kazma R, Babron MC, Genin E. Genetic association and gene-environment inter-
action: a new method for overcoming the lack of exposure information in con-
trols. Am J Epidemiol 2011; 173(2): 225–235.

Supplementary Information accompanies the paper on the Journal of Human Hypertension website (http://www.nature.com/jhh)

STK39 and hypertension
J Xu et al

297

& 2013 Macmillan Publishers Limited Journal of Human Hypertension (2013) 294 – 297

http://www.nature.com/jhh

	Lack of association between STK39 and hypertension in the Chinese population
	Introduction
	Methods
	Study population
	BP and clinical parameters
	SNP genotyping
	Statistics

	Results
	Discussion
	Acknowledgements
	Note
	References