Skip to main content
PMC home page
Pharmacogenomics logoLink to Pharmacogenomics
. 2018 Jan 10;19(3):175–184. doi: 10.2217/pgs-2017-0187

Attempted validation of 44 reported SNPs associated with tacrolimus troughs in a cohort of kidney allograft recipients

William S Oetting 1,1,*, Baolin Wu 2,2, David P Schladt 3,3, Weihua Guan 2,2, Rory P Remmel 4,4, Casey Dorr 3,3,5,5, Roslyn B Mannon 6,6, Arthur J Matas 7,7, Ajay K Israni 3,3,8,8,9,9, Pamala A Jacobson 1,1
PMCID: PMC6021962  PMID: 29318894

Abstract

Aim:

Multiple genetic variants have been associated with variation in tacrolimus (TAC) trough concentrations. Unfortunately, additional studies do not confirm these associations, leading one to question if a reported association is accurate and reliable. We attempted to validate 44 published variants associated with TAC trough concentrations.

Materials & methods:

Genotypes of the variants in our cohort of 1923 kidney allograft recipients were associated with TAC trough concentrations.

Results:

Only variants in CYP3A4 and CYP3A5 were significantly associated with variation in TAC trough concentrations in our validation.

Conclusion:

There is no evidence that common variants outside the CYP3A4 and CYP3A5 loci are associated with variation in TAC trough concentrations. In the future rare variants may be important and identified using DNA sequencing.

Keywords: : CYP3A5, GWAS, kidney, tacrolimus, transplantation

Obtaining optimal blood concentrations of the immunosuppressive agent tacrolimus (TAC) in kidney allograft recipients is critical to reducing the risk of acute rejection (AR) and maximizing graft survival, while reducing the risk of TAC-associated adverse effects [1]. Although individuals are given similar doses of TAC, there is significant variation in trough concentrations between individuals, caused in part by differences in TAC clearance [2]. Additionally, dose-adjusted trough concentrations differ significantly between populations. It is well known that African–American (AA) allograft kidney recipients have significantly lower TAC trough concentrations in whole blood when compared with European–American (EA) recipients at equivalent doses of TAC, and these lower concentrations have been associated with increased risk for AR [1,3–5]. Even within populations there is significant variability in TAC trough concentrations, putting some recipients at risk for either AR, due to low TAC concentrations, or adverse effects such as infections, nephrotoxicity, hypertension, hyperglycemia due to excessively high TAC concentrations.

To understand this variation in TAC trough concentrations, numerous studies in the last two decades have been undertaken to identify associated genetic variants in candidate SNPs thought to be involved in TAC pharmacokinetics (see Table 1) [6–56]. Unfortunately, there are additional studies which do not validate these associations. The lack of validation for the initial association is in many cases likely due to underpowered studies [57]. Most of these studies used small, single center cohorts with the majority of the studies having less than 100 recipients. In many studies, multiple genetic variants were analyzed without taking multiple testing into consideration, resulting in inflated type I errors with increased false positives. We and others have shown that variants in CYP3A4 and CYP3A5 are major contributors to this variation in TAC troughs and have been estimated to contribute to at least half of the observed variance when combined with clinical factors, leaving approximately half of the observed variance unexplained [32,54,58].

Table 1. . Variants identified from the literature previously associated with tacrolimus trough concentrations.

rs# Proxy SNP† Gene Chrom. Position Common name Nucleotide Amino acid Ref.
rs11265572   NR1I3 1 1.61E + 08   g.161243273G>T Noncoding [6]
rs1800872   IL10 1 2.07E + 08 -592C>A c.-627A>C Noncoding [7,8]
rs1800871   IL10 1 2.07E + 08 -819C>T c.-854T>C Noncoding [7,8]
rs1800896   IL10 1 2.07E + 08 -1082G>A c.-1117A>G Noncoding [9]
rs4553808 rs16840252 CTLA4 2 2.04E + 08   c.-1661A>G Noncoding [10]
rs3814055   NR1I2 3 1.2E + 08 -25385T>C c.-1135C>T Noncoding [11,12]
rs6785049   NR1I2 3 1.2E + 08 7635A/G c.795-93G>A Noncoding [13]
rs2276707 rs3814057 NR1I2 3 1.2E + 08 8055C>T c.938-17C>T Noncoding [14]
rs181781 rs657075 IL3 5 1.32E + 08   c.-1285G>A Noncoding [10]
rs1800796 rs1524107 IL6 7 22726627   c.-636G>C Noncoding [15]
rs1057868   POR 7 75985688 *28 c.1508C>T p.Ala503Val [16–22]
rs1045642 rs2235048 ABCB1 7 87509329 3435C/T c.3435T>C p.Ile1145= [12,23–29]
rs2032582 rs4148738 ABCB1 7 87531302 2677G/T/A c.2677T/A/G p.Ser893Ala/Thr [12,24,27–30]
rs1128503   ABCB1 7 87550285 1236C/T c.1236T>C p.Gly412= [12,24,28–30]
rs2229109 rs117937072 ABCB1 7 87550493 1199G/A c.1199G>A p.Ser400Asn [30,31]
rs9282564   ABCB1 7 87600124 61A/G c.61A>G p.Asn21Asp [24]
rs3213619   ABCB1 7 87600877 -129T/C c.-129T>C Noncoding [13]
rs41303343   CYP3A5 7 99652770 *7 c.1035_1036insT p.Thr346Tyrfs [32,33]
rs10264272   CYP3A5 7 99665212 *6 c.624G>A p.Lys208= [32,33]
rs4646450   CYP3A5 7 99668695   c.319-1630C>T Noncoding [34]
rs776746   CYP3A5 7 99672916 *3 c.219-237A>G Noncoding [10–12,16,18,19,23,25,26,29,30,32,34–42]
rs2257401   CYP3A7 7 99709062 *2 c.1226C>G p.Thr409Arg [11]
rs2242480   CYP3A4 7 99763843 *1G c.1023 + 12G>A Noncoding [7,10–12,29]
rs28371759   CYP3A4 7 99764003 *18B c.875T>C p.Leu292Pro [43,44]
rs4646437   CYP3A4 7 99767460   c.671-205C>T Noncoding [7,10]
rs35599367   CYP3A4 7 99768693 *22 c.522-191C>T Noncoding [16,17,20,21,39,45–47]
rs2740574   CYP3A4 7 99784473 *1B c.-392G>A Noncoding [42,48–51]
rs1927907   TLR4 9 1.18E + 08   c.140 + 1757C>T Noncoding [52]
rs4986893   CYP2C19 10 94780653 *3 c.636G>A p.Trp212Ter [40]
rs4244285 rs12571421 CYP2C19 10 94781859 *2 c.681G>A p.Pro227= [37,40,53]
rs2273697   ABCC2 10 99804058 1249G>A c.1249G>A p.Val417Ile [42]
rs3740066   ABCC2 10 99844450 3972C>T c.3972C>T p.Ile1324= [42]
rs2070673   CYP2E1 10 1.34E + 08   c.-333A>T Noncoding [37]
rs2237991   ABCC8 11 17418682   c.2223-1720T>C Noncoding [54]
rs5744247   IL18 11 1.12E + 08   c.-8-372C>G Noncoding [38]
rs1946518 rs1946519 IL18 11 1.12E + 08 A-607C c.-838A>C Noncoding [41]
rs4149117   SLCO1B3 12 20858546 T334G c.335C>A p.Ser112Ala [55]
rs7311358   SLCO1B3 12 20862826 G699A c.699G>A p.Met233Ile [55]
rs2306283   SLCO2B1 12 21176804 A388G c.388A>G p.Asn130Asp [30]
rs4149056   SLCO2B1 12 21178615 T521C c.521T>C p.Val174Ala [30]
rs3745274   CYP2B6 19 41006936 *6 c.516G>T p.Gln172His [37]
rs2239393   COMT 22 19962905   c.289 + 90A>G Noncoding [56]
rs4823613 rs4253730 PPARA 22 46202410   c.208 + 3819A>G Noncoding [19]
rs4253728 rs4253730 PPARA 22 46214170   c.209-1003G>A Noncoding [19,46]
Open in a new tab

†SNP was used as a proxy when a variant was not present in the genotyping chip.

We attempted to validate 44 variants previously reported in the literature to be associated with variation in TAC troughs using DNA from a cohort of kidney allograft recipients on which a genome wide association study (GWAS) had been performed and clinical information from kidney recipients enrolled in the long-term Deterioration of Kidney Allograft Function (DeKAF) Genomics study. This cohort consists of 1923 kidney recipients with 31,906 TAC trough concentrations and doses.

Materials & methods

The design of the DeKAF Genomics study and cohort characteristics has been previously reported [32,59,60]. For this analysis, kidney transplant recipients with genome wide association study data were identified and divided into two sub-cohorts consisting of 1560 EA and 363 AA kidney allograft recipients and tested separately (Table 2). Though, self-reported race was available in the clinical information, subjects were separated into EA and AA cohorts using principal components using the GWAS. Subjects were aged 18 years and older, received TAC for maintenance immunosuppression and had TAC troughs and doses available in the first 6 months post-transplant. Subjects were enrolled at time of transplant and signed informed consents were approved by the Institutional Review Boards of the enrolling centers. This study is registered at www.clinicaltrials.gov (NCT01714440).

Table 2. . Recipient characteristics for African–American and European–American recipients.

Characteristics European–American (n = 1560) African–American (n = 363) p-value
Age group (years), n (%)     <0.0001
– 18–34 180 (11.5%) 71 (19.6%)  
– 35–64 1121 (71.9%) 272 (74.9%)  
– 65–84 259 (16.6%) 20 (5.5%)  
Donor age group (years), n (%)     <0.0001
– 0–34 485 (31.1%) 171 (47.1%)  
– 35–64 1031 (66.1%) 184 (50.7%)  
– 65–84 44 (2.8%) 8 (2.2%)  
Living donor status, n (%) 1031 (66.1%) 112 (30.9%) <0.0001
Female, n (%) 579 (37.1%) 132 (36.4%) 0.81
Diabetes at transplant, n (%) 611 (39.2%) 133 (36.6%) 0.40
SPK, n (%) 130 (8.3%) 18 (5.0%) 0.029
Body mass index, mean (SD) 28.3 (5.5) 28.8 (5.4) 0.075
Antibody induction, n (%)     <0.0001
– Monoclonal 595 (38.1%) 176 (48.5%)  
– Polyclonal 857 (54.9%) 175 (48.2%)  
– None 63 (4.0%) 6 (1.7%)  
– Combination 45 (2.9%) 6 (1.7%)  
Median tacrolimus trough (IQR) in the first 34 days in ng/ml 8.7 (6.6–10.8) 5.4 (3.5–7.9) <0.0001
Median tacrolimus trough (IQR) after day 34 in ng/ml 8.1 (6.5–9.9) 7.1 (5.4–9.0) <0.0001
Median daily tacrolimus dose (IQR) in first 24 days in mg 6.0 (4.0–8.0) 6.0 (4.0–8.0) 0.77
Median daily tacrolimus dose (IQR) after day 25 in mg 5.0 (3.5–8.0) 8.0 (6.0–10.0) <0.0001
Median dose-normalized tacrolimus trough (IQR) in ng/ml per total daily dose in mg 1.52 (1.00–2.33) 0.78 (0.52–1.22) <0.0001
Open in a new tab

IQR: Interquartile range; SD: Standard deviation.

Clinical information was obtained through the DeKAF Genomics study and obtained from the respective medical records [59,60]. Participants received oral immediate release TAC therapy with mycophenolate maintenance with varying durations of steroid per transplant center standard-of-care protocols. Induction therapy was administered as per transplant center preference but mainly consisted of rabbit antithymocyte globulin, basiliximab or Campath-1H. Immunologically, high risk patients were more likely to receive rabbit antithymocyte globulin, such as those with donor specific antibody, pregnancies or repeat transplants. TAC troughs were clinically measured at each site and were analyzed in a clinical laboratory improvement amendments approved laboratory and >95% were measured from whole blood by liquid chromatography-mass spectrometry. When available, two measurements were obtained in the first 8 weeks, and two levels per month in months 3, 4, 5 and 6 for a maximum of 24 trough concentrations per patient. TAC doses were adjusted based on trough concentrations to reach institution-specific trough goals based on time post-transplant (generally 8–12 ng/ml in months 0–3 and 6–10 ng/ml in months 4–6). Doses were also adjusted for toxicity (e.g., nephrotoxicity) by center specific preferences. Trough values were normalized for dose (ng/ml per total daily dose in mg) prior to statistical analysis.

An extensive review of the literature was conducted using PubMed and published variants that were reported to be significantly associated with variation in TAC trough concentrations in solid organ transplantation patients were identified. Only variants which were shown to have a statistically significant association (p < 0.05 as stated in the published report) were included in this study. Most studies analyzed TAC concentrations in kidney recipients. The variants which were chosen from the literature for validation in this report are shown in Table 1. Genotype information for this study was extracted from our previous study using an Affymetrix TxArray GWAS chip created specifically for analysis of allograft recipients [32,61]. Polymorphisms capturing Ancestry informative markers and 7500 drug adsorption, metabolism, excretion and toxicity markers (ADME; n = approximately 7500) including SNPs from PharmGKB were added to this chip [61,63]. Also included were additional SNPs extracted from the Affymetrix-Biobank to increase coverage of African and other populations [61]. A total of 644,224 SNPs were available for genotyping data. A total of 44 variants in 22 genes were identified in the literature and analyzed using the genotypes from this chip for each individual. For those variants which were not part of the GWAS chip, a proxy SNP was selected which was present on the chip and genotypes available. In all cases, the r2 between the reported variant and the proxy SNP was 1.0 as determined by the SNP Annotation and Proxy Search program (SNAP; 62), with the exception of rs181781 and rs2276707 (r2 = 0.938), rs4823613 and rs4253730 (r2 = 0.959) and rs4253728 and rs4253730 (r2 = 0.920).

Statistical analysis for validation of TAC trough concentration associated variants

A linear mixed effects model was used to test for the association between genotypes of each SNP and the longitudinal dose-normalized TAC trough concentrations in each cohort. To achieve a better normality approximation, we log transformed the dose-normalized TAC concentrations. A spline model with a change of slope at day 9 post-transplant was used to model the varying time effect of trough concentrations following previous approaches [32,56]. We included a random intercept and slope for days post-transplant and modeled the additive effects of genotypes, adjusting for age, gender and transplant center. To adjust for potential population stratification, the first ancestry principal component was incorporated in the regression models instead of their reported race. The threshold for significance was set at p < 0.001 after taking multiple testing into consideration (n = 44 tests).

For comparison between the two cohorts, Fisher Exact test was used for categorical variables and t-test for continuous variables. TAC total daily dose-normalized trough doses and troughs were compared using simple linear mixed effects longitudinal models with an effect for days post-transplant and a random intercept in each cohort. The data were visually inspected to determine the point of divergence post-transplant of troughs (day 24) and doses (day 34) and used for the time points for comparison between the two cohorts.

Results

Characteristics of the two cohorts are shown in Table 2. Significant differences between the EA and AA cohorts included recipient age (p < 0.0001) where recipients in the EA cohort were older, donor age (p < 0.0001) where the EA recipients received older donor allografts, living donor status (p < 0.0001) where EA recipients had a higher percentage of living donors, and type of antibody induction (p < 0.0001). Compared with AA recipients, the median TAC trough was significantly higher in EA recipients p < 0.0001) in the first 34 days post-transplant as was the median dose-normalized TAC trough (p < 0.0001). Median daily TAC doses were similar in the two cohorts until day 24 when after day 25, the doses were significantly higher in AA subjects (p < 0.0001) compared with the EA subjects.

The significance of association for each published variant tested for validation in the DeKAF cohort is shown in Table 3. We were able to validate seven of the previously reported variants in the EA cohort and all were in the CYP3A locus region (p < 0.001). All other variants were below the threshold of significance. All significant variants, with the exception of CYP3A4*22 (rs35599367), were found to be in linkage disequilibrium (LD) with rs776746 (CYP3A5*3), using the CEU population panel for LD testing (Table 4). CEU is defined by 1000 genomes as Utah Residents (CEPH) with northern and western European ancestry.

Table 3. . Validation of published variants for tacrolimus trough concentrations in European–American and African–American cohorts in the Deterioration of Kidney Allograft Function genomics study.

rs_id Gene Variant EA AA EA AA  
      Freq.† Freq.† Beta p-value Beta p-value
rs11265572 NR1I3 g.161243273G>T 0.001 0.004 NA NA NA NA
rs1800872 IL10 -592C>A 0.236 0.398 -0.039 0.066 -0.014 0.735
rs1800871 IL10 -819C>T 0.236 0.398 -0.039 0.066 -0.014 0.735
rs1800896 IL10 -1082G>A 0.498 0.356 0.022 0.218 0.016 0.707
rs16840252‡ CTLA4   0.191 0.171 0.01 0.655 -0.113 0.032
rs3814055 NR1I2 -25385T>C 0.383 0.283 -0.017 0.352 -0.026 0.558
rs6785049 NR1I2 7635A/G 0.389 0.095 0.023 0.218 0.06 0.384
rs3814057‡ NR1I2 8055C>T 0.19 0.441 0.016 0.478 0.03 0.467
rs657075‡ IL3   0.099 0.031 0.054 0.075 0.027 0.802
rs1524107‡ IL6 c.-636G>C 0.052 0.066 -0.074 0.075 -0.035 0.656
rs1057868 POR *28 0.277 0.201 0.009 0.657 0.044 0.375
rs2235048‡ ABCB1 3435C/T 0.458 0.202 0.006 0.758 0.075 0.134
rs4148738‡ ABCB1 2677G/T/A 0.442 0.235 0.01 0.587 0.062 0.201
rs117937072‡ ABCB1 1199G/A 0.041 0.009 -0.01 0.817 NA NA
rs1128503 ABCB1 1236C/T 0.427 0.202 0.014 0.455 0.066 0.19
rs9282564 ABCB1 61A/G 0.108 0.004 0.001 0.968 NA NA
rs3213619 ABCB1 -129T/C 0.036 0.076 -0.036 0.462 -0.053 0.466
rs41303343 CYP3A5 *7 0 0.109 NA NA 0.361 1.18E-09
rs10264272 CYP3A5 *6 0.001 0.123 NA NA 0.067 0.269
rs4646450 CYP3A5   0.156 0.133 -0.305 3.46E-38 0.271 4.91E-07
rs776746 CYP3A5 *3 0.068 0.29 -0.657 6.08E-98 0.299 2.34E-14
rs2257401 CYP3A7 *2 0.087 0.46 -0.538 1.46E-79 0.094 0.015
rs2242480 CYP3A4 *1G 0.094 0.254 -0.487 8.55E-68 0.236 1.91E-08
rs28371759 CYP3A4 *18B 0 0.006 NA NA NA NA
rs4646437 CYP3A4   0.11 0.263 -0.405 5.63E-53 0.223 8.86E-08
rs35599367 CYP3A4 *22 0.057 0.004 0.339 4.81E-19 NA NA
rs2740574 CYP3A4 *1B 0.037 0.346 -0.464 1.74E-24 0.026 0.502
rs1927907 TLR4 c.140 + 1757C>T 0.13 0.226 0.034 0.209 -0.064 0.172
rs4986893 CYP2C19 *3 0 0.001 NA NA NA NA
rs12571421‡ CYP2C19 *2 0.153 0.177 0.02 0.428 0.026 0.603
rs2273697 ABCC2 1249G>A 0.214 0.153 0.023 0.286 0.024 0.656
rs3740066 ABCC2 3972C>T 0.359 0.257 -0.018 0.331 -0.057 0.197
rs2070673 CYP2E1 c.-333A>T 0.145 0.315 0.012 0.631 -0.032 0.46
rs2237991 ABCC8 c.2223–1720T>C 0.252 0.332 0.027 0.195 -0.039 0.349
rs5744247 IL18 c.-8–372C>G 0.108 0.023 -0.002 0.933 -0.171 0.197
rs1946519‡ IL18 c.-838A>C 0.399 0.332 -0.003 0.861 -0.038 0.378
rs4149117 SLCO1B3 T334G 0.147 0.387 -0.014 0.591 0.093 0.022
rs7311358 SLCO1B3 G699A 0.145 0.393 -0.012 0.632 0.084 0.037
rs2306283 SLCO1B1 A388G 0.406 0.206 -0.007 0.708 0 0.994
rs4149056 SLCO1B1 T521C 0.156 0.024 -0.022 0.371 -0.052 0.684
rs3745274 CYP2B6 *6 0.247 0.374 -0.001 0.945 -0.014 0.732
rs2239393 COMT c.289 + 90A>G 0.39 0.409 -0.018 0.326 0.005 0.891
rs4253730‡ PPARA c.209–1003G>A 0.275 0.47 0.015 0.44 0.027 0.499
Open in a new tab

†MAF based on the cohort genotypes. Minor alleles are those which differ from the reference sequence.

‡ Proxy variant used (see Table 1).

AA: African–American; EA: European–American; MAF: Minor allele frequency; NA: Not analyzed due to low minor allele frequency.

Table 4. . Linkage disequilibrium between significant SNPs in the CYP3A locus and rs776746 (CYP3A5*3) in European–American (CEU) and African (YRI) populations.

rs# Gene Variant r2 (CEU) r2 (YRI)
rs776746 CYP3A5 *3 1 1
rs2257401 CYP3A7 *2 0.87 0.34
rs2242480 CYP3A4 *1G 0.68 0.39
rs4646437 CYP3A4   0.56 0.27
rs4646450 CYP3A5   0.28 0
rs2740574 CYP3A4 *1B 0.41 0.1
rs35599367 CYP3A4 *22 0.03 0
Open in a new tab

For the AA cohort, we were able to validate five variants (Table 3) and as in the EA cohort, only variants in the CYP3A locus region were statistically significant. Out of these five variants, two were in LD with rs776746 using the YRI population panel for LD testing, CYP3A4*1G (rs2242480; r2 = 0.387) and rs4646437 (r2 = 0.273). The LoF variant CYP3A5*7 (rs41303343) was not in LD with CYP3A5*3. The variant rs4646450 was not in LD with rs776746 in the YRI population panel, but was when the CEU panel was used. YRI is defined by 1000 genomes as Yoruba in Ibadan, Nigeria. The LoF variant CYP3A5*6 was not statistically significant in this analysis, but is likely due to the low number of individuals with this variant in the AA cohort.

There were also a few variants that presented with suggestive significance. In the EA cohort four variants in three interleukin genes: IL10 (rs1800871; 0.066 and rs1800872; p = 0.066), IL3 (rs181781; p = 0.075) and IL6 (rs1800796; p = 0.075) had p-values below 0.1. The two IL10 variants are in complete LD (r2 = 1.000). In the AA cohort, two variants in the solute carrier organic anion transporter family member 1B3 (SLCO1B3) presented with suggestive significance (rs4149117; p = 0.022 and rs7311358; p = 0.037). These two variants in SLCO1B3 are in complete LD (r2 = 1.000). Additionally, a variant in the CTLA4 gene (rs16840252, a proxy for rs4553808, was also suggestive (p = 0.032). In all of these variants, significance was below the threshold after multiple testing is taken into consideration.

Discussion

Optimizing TAC blood concentrations is critical to maximizing graft survival for kidney allograft recipients. Unfortunately, trough levels can vary widely between recipients, even when taking the same dose. One reason for this variation is difference in the pharmacokinetics of TAC between individuals in part due to genetic variation in critical TAC metabolizing enzymes. Identification of these genetic variants would help in explaining this variation in drug concentrations and provide a tool to personalize dosing.

Numerous studies have reported genetic variants with a statistically significant association with TAC trough concentrations (Table 1). The genes containing these variants include drug metabolism enzymes (e.g., CYP3A4 and CYP3A5), drug transporters (e.g., ABCB1 and SLCO1B3), transcriptional factors which effect CYP3A4 expression (e.g., NR1I2 and NR1I3) and members of the interleukin family of genes (e.g., IL3 and IL6) as well as others. In this analysis, only LoF alleles in CYP3A5, and the CYP3A4*22 allele in the EA cohort, were found to be statistically significant for association with TAC troughs. Some of the variants we tested were shown to be statistically significant in numerous previous reports including POR*28 (rs1057868) and ABCB1 variants c.3435T>C (rs1045642), c.2677T/A/G (rs2032582) and c.1236T>C (rs1128503). Though, subsequent reports appear to validate the original findings for these four variants, our larger cohort did not identify a significant association with these variants

In this analysis, only variants within the CYP3A4 and CPY3A5 genes were identified, and most of those variants were in LD with a known functional variant leaving only three variants which were associated with variation in TAC trough levels in this study, CYP3A5*3, CYP3A5*7 and CYP3A4*22. This is not surprising because the CYP3A4 and CYP3A5 enzymes are the major metabolizing enzymes for TAC.

There are several reasons for the lack of validation of the majority of the variants tested. For the most part, the original reports utilized small underpowered cohorts for the initial association. Also, most studies did not follow-up with an additional cohort to confirm their findings. Additional reasons for the lack of validation include possible center effects when recipients are from multiple centers. Clinical differences in practice and concomitant medications that may influence clearance, age of the recipient, time post-transplant, adherence, use of self-reported race instead of principal components analysis resulting in a combination of racial groups are additional sources of error.

This study focused on common (high minor allele frequency) variants. We have shown that only variants within the CYP3A4 and CYP3A5 genes, along with clinical factors, account for approximately 40% of the variance [58]. This leaves a significant percentage of unexplained variance associated with TAC trough concentrations [58]. Some of this missing unexplained variance may be due to additional genetic variance (missing heritability) or it may be a result of an over estimation of the heritability which exists. This analysis may also lack additional clinical variables not adjusted for, such as non-adherence. Any additional heritability which still exists most likely is not due to variants with a high minor allele frequency (MAF>0.01). We hypothesize that the source of any missing heritability will be low allele frequency variants in genes influencing TAC pharmacokinetics. DNA sequencing of these genes will reveal if these variants exist and these studies are on-going in our cohort.

Conclusion

Validation of identified variants associated with a specific outcome is necessary before this information can be translated into clinical care. A large number of genetic variants have been previously reported to be associated with variability in TAC trough concentrations. Unfortunately, most of these studies were underpowered. In our large cohort of kidney allograft recipients, we found that only functional variants in CYP3A4 and CYP3A5 could be validated.

Future perspective

Optimized serum concentrations of TAC in allograft recipients are critical for graft survival. Identifying which variants are important for individualized immunosuppression will allow for the development of dosing equations utilizing the proper genomic data. In this report, we were able to validate those common genetic variants which are associated with TAC trough concentrations, while eliminating those that have been previously reported as associated with this variation. We are using genetic information, along with clinical variables, to optimize dosing equations for TAC, to better estimate the initial dose for kidney allograft recipients.

Summary points.

  • We attempted to validate 44 common variants in 22 loci previously reported to be associated with variation in tacrolimus trough concentrations.

  • We used DNA samples, clinical information and tacrolimus trough concentrations from 1560 EA and 363 AA kidney allograft recipients.

  • Common functional variants in CYP3A4 and CYP3A5 were associated with variation in tacrolimus trough concentrations.

  • All other variants tested were not validated.

  • This information may improve the precision of genotype-guided dosing.

Acknowledgements

The authors wish to thank the research subjects for their participation in this study. We acknowledge the dedication and hard work of our coordinators at each of the DeKAF Genomics clinical sites: University of Alberta, N Bobocea, T Wong, A Geambasu and A Sader; University of Manitoba, M Ross and K Peters; University of Minnesota, M DeGrote, M Myers and D Berglund; Hennepin County Medical Center, L Berndt; Mayo Clinic, T DeLeeuw; University of Iowa, W Wallace and T Lowe; University of Alabama, J Vaughn, V Stephens and T Hilario. We also acknowledge the dedicated work of our research scientists M Brott and A Muthusamy.

Footnotes

Financial & competing interests disclosure

This study was supported in part by NIH/NIAID grants 5U19-AI070119 and 5U01-AI058013. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

References

Papers of special note have been highlighted as: • of interest; •• of considerable interest

  • 1.Arreola-Guerra JM, Serrano M, Morales-Buenrostro LE, Vilatobá M, Alberú J. Tacrolimus trough levels as a risk factor for acute rejection in renal transplant patients. Ann. Transplant. 2016;21:105–114. doi: 10.12659/aot.895104. [DOI] [PubMed] [Google Scholar]
  • 2.Egeland EJ, Robertsen I, Hermann M, et al. High tacrolimus clearance is a risk factor for acute rejection in the early phase after renal transplantation. Transplantation. 2017;101(8):e273–e279. doi: 10.1097/TP.0000000000001796. [DOI] [PubMed] [Google Scholar]
  • 3.Neylan JF. Effect of race and immunosuppression in renal transplantation: three-year survival results from a US multicenter, randomized trial. FK506 Kidney Transplant Study Group. Transplant Proc. 1998;30(4):1355–1358. doi: 10.1016/s0041-1345(98)00274-7. [DOI] [PubMed] [Google Scholar]
  • 4.Jacobson PA, Oetting WS, Brearley AM, et al. Novel polymorphisms associated with tacrolimus trough concentrations: results from a multicenter kidney transplant consortium. Transplantation. 2011;91(3):300–308. doi: 10.1097/TP.0b013e318200e991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Israni AK, Riad SM, Leduc R, et al. Tacrolimus trough levels after month 3 as a predictor of acute rejection following kidney transplantation: a lesson learned from DeKAF genomics. Transpl. Int. 2013;26(10):982–989. doi: 10.1111/tri.12155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Chen D, Guo F, Shi J, et al. Association of hemoglobin levels, CYP3A5, and NR1I3 gene polymorphisms with tacrolimus pharmacokinetics in liver transplant patients. Drug Metab. Pharmacokinet. 2014;29(3):249–253. doi: 10.2133/dmpk.dmpk-13-rg-095. [DOI] [PubMed] [Google Scholar]
  • 7.Li CJ, Li L, Lin L, et al. Impact of the CYP3A5, CYP3A4, COMT, IL-10 and POR genetic polymorphisms on tacrolimus metabolism in Chinese renal transplant recipients. PLoS ONE. 2014;9(1):e86206. doi: 10.1371/journal.pone.0086206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zhang X, Wang Z, Fan J, Liu G, Peng Z. Impact of interleukin-10 gene polymorphisms on tacrolimus dosing requirements in Chinese liver transplant patients during the early posttransplantation period. Eur. J. Clin. Pharmacology. 2011;67(8):803–813. doi: 10.1007/s00228-011-0993-8. [DOI] [PubMed] [Google Scholar]
  • 9.Li D, Zhu JY, Wang X, Lou YQ, Zhang GL. Polymorphisms of tumor necrosis factor-alpha, interleukin-10, cytochrome P450 3A5 and ABCB1 in Chinese liver transplant patients treated with immunosuppressant tacrolimus. Clin. Chim. Acta. 2007;383(1–2):133–139. doi: 10.1016/j.cca.2007.05.008. [DOI] [PubMed] [Google Scholar]
  • 10.Liu MZ, He HY, Zhang YL, et al. IL-3 and CTLA4 gene polymorphisms may influence the tacrolimus dose requirement in Chinese kidney transplant recipients. Acta Pharmacol. Sin. 2017;38(3):415–423. doi: 10.1038/aps.2016.153. 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Choi Y, Jiang F, An H, Park HJ, Choi JH, Lee H. A pharmacogenomic study on the pharmacokinetics of tacrolimus in healthy subjects using the DMET™ Plus platform. Pharmacogenomics J. 2016;17(2):174–179. doi: 10.1038/tpj.2015.99. [DOI] [PubMed] [Google Scholar]
  • 12.Li JI, Liu S, Fu Q, et al. Interactive effects of CYP3A4, CYP3A5, MDR1 and NR1I2 polymorphisms on tracrolimus trough concentrations in early postrenal transplant recipients. Pharmacogenomics. 2015;16(12):1355–1365. doi: 10.2217/pgs.15.78. [DOI] [PubMed] [Google Scholar]
  • 13.Press RR, Ploeger BA, den Hartigh J, et al. Explaining variability in tacrolimus pharmacokinetics to optimize early exposure in adult kidney transplant recipients. Ther. Drug Monit. 2009;31(2):187–197. doi: 10.1097/FTD.0b013e31819c3d6d. [DOI] [PubMed] [Google Scholar]
  • 14.Barraclough KA, Isbel NM, Lee KJ, et al. NR1I2 polymorphisms are related to tacrolimus dose-adjusted exposure and BK viremia in adult kidney transplantation. Transplantation. 2012;94(10):1025–1032. doi: 10.1097/TP.0b013e31826c3985. [DOI] [PubMed] [Google Scholar]
  • 15.Chen D, Fan J, Guo F, Qin S, Wang Z, Peng Z. Novel single nucleotide polymorphisms in interleukin 6 affect tacrolimus metabolism in liver transplant patients. PLoS ONE. 2013;8(8):e73405. doi: 10.1371/journal.pone.0073405. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  • 16.Elens L, Hesselink DA, Bouamar R, et al. Impact of POR*28 on the pharmacokinetics of tacrolimus and cyclosporine A in renal transplant patients. Ther. Drug Monit. 2014;36(1):71–79. doi: 10.1097/FTD.0b013e31829da6dd. [DOI] [PubMed] [Google Scholar]
  • 17.Hesselink DA, Bouamar R, Elens L, van Schaik RH, van Gelder T. The role of pharmacogenetics in the disposition of and response to tacrolimus in solid organ transplantation. Clin. Pharmacokinet. 2014;53(2):123–139. doi: 10.1007/s40262-013-0120-3. [DOI] [PubMed] [Google Scholar]
  • 18.Lesche D, Sigurdardottir V, Setoud R, et al. CYP3A5*3 and POR*28 genetic variants influence the required dose of tacrolimus in heart transplant recipients. Ther. Drug Monit. 2014;36(6):710–715. doi: 10.1097/FTD.0000000000000080. [DOI] [PubMed] [Google Scholar]
  • 19.Lunde I, Bremer S, Midtvedt K, et al. The influence of CYP3A, PPARA, and POR genetic variants on the pharmacokinetics of tacrolimus and cyclosporine in renal transplant recipients. Eur. J. Clin. Pharmacol. 2014;70(6):685–693. doi: 10.1007/s00228-014-1656-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Tang JT, Andrews LM, van Gelder T, et al. Pharmacogenetic aspects of the use of tacrolimus in renal transplantation: recent developments and ethnic considerations. Expert Opin. Drug Metab. Toxicol. 2016;12(5):555–565. doi: 10.1517/17425255.2016.1170808. [DOI] [PubMed] [Google Scholar]
  • 21.Kuypers DR, de Loor H, Naesens M, Coopmans T, de Jonge H. Combined effects of CYP3A5*1, POR*28, and CYP3A4*22 single nucleotide polymorphisms on early concentration-controlled tacrolimus exposure in de-novo renal recipients. Pharmacogenet. Genomics. 2014;24(12):597–606. doi: 10.1097/FPC.0000000000000095. [DOI] [PubMed] [Google Scholar]
  • 22.Jannot AS, Vuillemin X, Etienne I, et al. A lack of significant effect of POR*28 allelic variant on tacrolimus exposure in kidney transplant recipients. Ther. Drug Monit. 2016;38(2):223–229. doi: 10.1097/FTD.0000000000000267. [DOI] [PubMed] [Google Scholar]
  • 23.Yan L, Li Y, Tang JT, An YF, Wang LL, Shi YY. Donor ABCB1 3435 C>T genetic polymorphisms influence early renal function in kidney transplant recipients treated with tacrolimus. Pharmacogenomics. 2016;17(3):249–257. doi: 10.2217/pgs.15.165. [DOI] [PubMed] [Google Scholar]
  • 24.Ruiz J, Herrero MJ, Bosó V, et al. Impact of single nucleotide polymorphisms (SNPs) on immunosuppressive therapy in lung transplantation. Int. J. Mol. Sci. 2015;16(9):20168–20182. doi: 10.3390/ijms160920168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Naito T, Mino Y, Aoki Y, et al. ABCB1 genetic variant and its associated tacrolimus pharmacokinetics affect renal function in patients with rheumatoid arthritis. Clin. Chim. Acta. 2015;445:79–84. doi: 10.1016/j.cca.2015.03.021. [DOI] [PubMed] [Google Scholar]
  • 26.Knops N, van den Heuvel LP, Masereeuw R, et al. The functional implications of common genetic variation in CYP3A5 and ABCB1 in human proximal tubule cells. Mol. Pharm. 2015;12(3):758–768. doi: 10.1021/mp500590s. [DOI] [PubMed] [Google Scholar]
  • 27.Kravljaca M, Perovic V, Pravica V, et al. The importance of MDR1 gene polymorphisms for tacrolimus dosage. Eur. J. Pharm. Sci. 2016;83:109–113. doi: 10.1016/j.ejps.2015.12.020. [DOI] [PubMed] [Google Scholar]
  • 28.Cusinato DA, Lacchini R, Romao EA, Moysés-Neto M, Coelho EB. Relationship of CYP3A5 genotype and ABCB1 diplotype to tacrolimus disposition in Brazilian kidney transplant patients. Br. J. Clin. Pharmacol. 2014;78(2):364–372. doi: 10.1111/bcp.12345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Genvigir FD, Salgado PC, Felipe CR, et al. Influence of the CYP3A4/5 genetic score and ABCB1 polymorphisms on tacrolimus exposure and renal function in Brazilian kidney transplant patients. Pharmacogenet. Genomics. 2016;26(10):462–472. doi: 10.1097/FPC.0000000000000237. [DOI] [PubMed] [Google Scholar]
  • 30.Elens L, Capron A, Kerckhove VV, et al. 1199G>A and 2677G>T/A polymorphisms of ABCB1 independently affect tacrolimus concentration in hepatic tissue after liver transplantation. Pharmacogenet. Genomics. 2007;17(10):873–883. doi: 10.1097/FPC.0b013e3282e9a533. [DOI] [PubMed] [Google Scholar]
  • 31.Dessilly G, Elens L, Panin N, et al. ABCB1 1199G>A genetic polymorphism (rs2229109) influences the intracellular accumulation of tacrolimus in HEK293 and K562 recombinant cell lines. PLoS ONE. 2014;9(3):e91555. doi: 10.1371/journal.pone.0091555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Oetting WS, Schladt DP, Guan W, et al. Genome wide association study of tacrolimus concentrations in African–American kidney transplant recipients identifies multiple CYP3A5 ALLELES. Am. J. Transplant. 2016;16(2):574–582. doi: 10.1111/ajt.13495. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Santoro A, Felipe CR, Tedesco-Silva H, et al. Pharmacogenetics of calcineurin inhibitors in Brazilian renal transplant patients. Pharmacogenomics. 2011;12(9):1293–1303. doi: 10.2217/pgs.11.70. [DOI] [PubMed] [Google Scholar]
  • 34.Onizuka M, Kunii N, Toyosaki M, et al. Cytochrome P450 genetic polymorphisms influence the serum concentration of calcineurin inhibitors in allogeneic hematopoietic SCT recipients. Bone Marrow Transplant. 2011;46:1113–1117. doi: 10.1038/bmt.2010.273. [DOI] [PubMed] [Google Scholar]
  • 35.Mac Guad R, Zaharan NL, Chik Z, Mohamed Z, Peng NK, Adnan WA. Effects of CYP3A5 genetic polymorphism on the pharmacokinetics of tacrolimus in renal transplant recipients. Transplant Proc. 2016;48(1):81–87. doi: 10.1016/j.transproceed.2016.01.001. [DOI] [PubMed] [Google Scholar]
  • 36.Nair SS, Sarasamma S, Gracious N, George J, Anish TS, Radhakrishnan R. Polymorphism of the CYP3A5 gene and its effect on tacrolimus blood level. Exp. Clin. Transplant. 2015;(Suppl. 1):197–200. [PubMed] [Google Scholar]
  • 37.Khaled SK, Palmer JM, Herzog J, et al. Influence of absorption, distribution, metabolism, and excretion genomic variants on tacrolimus/sirolimus blood levels and graft-versus-host disease after allogeneic hematopoietic cell transplantation. Biol. Blood Marrow Transplant. 2016;22(2):268–276. doi: 10.1016/j.bbmt.2015.08.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Fan J, Zhang X, Ren L, et al. Donor IL-18 rs5744247 polymorphism as a new biomarker of tacrolimus elimination in Chinese liver transplant patients during the early post-transplantation period: results from two cohort studies. Pharmacogenomics. 2015;16(3):239–250. doi: 10.2217/pgs.14.166. [DOI] [PubMed] [Google Scholar]
  • 39.de Jonge H, Elens L, de Loor H, van Schaik RH, Kuypers DR. The CYP3A4*22 C>T single nucleotide polymorphism is associated with reduced midazolam and tacrolimus clearance in stable renal allograft recipients. Pharmacogenomics J. 2015;15(2):144–152. doi: 10.1038/tpj.2014.49. [DOI] [PubMed] [Google Scholar]
  • 40.Iwamoto T, Monma F, Fujieda A, et al. Effect of genetic polymorphism of CYP3A5 and CYP2C19 and concomitant use of voriconazole on blood tacrolimus concentration in patients receiving hematopoietic stem cell transplantation. Ther. Drug Monit. 2015;37(5):581–588. doi: 10.1097/FTD.0000000000000182. [DOI] [PubMed] [Google Scholar]
  • 41.Xing J, Zhang X, Fan J, Shen B, Men T, Wang J. Association between interleukin-18 promoter variants and tacrolimus pharmacokinetics in Chinese renal transplant patients. Eur. J. Clin. Pharmacol. 2015;71(2):191–198. doi: 10.1007/s00228-014-1785-8. [DOI] [PubMed] [Google Scholar]
  • 42.Ogasawara K, Chitnis SD, Gohh RY, Christians U, Akhlaghi F. Multidrug resistance-associated protein 2 (MRP2/ABCC2) haplotypes significantly affect the pharmacokinetics of tacrolimus in kidney transplant recipients. Clin. Pharmacokinet. 2013;52(9):751–762. doi: 10.1007/s40262-013-0069-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Li DY, Teng RC, Zhu HJ, Fang Y. CYP3A4/5 polymorphisms affect the blood level of cyclosporine and tacrolimus in Chinese renal transplant recipients. Int. J. Clin. Pharmacol. Ther. 2013;51(6):466–474. doi: 10.5414/CP201836. [DOI] [PubMed] [Google Scholar]
  • 44.Shi XJ, Geng F, Jiao Z, Cui XY, Qiu XY, Zhong MK. Association of ABCB1, CYP3A4*18B and CYP3A5*3 genotypes with the pharmacokinetics of tacrolimus in healthy Chinese subjects: a population pharmacokinetic analysis. J. Clin. Pharm. Ther. 2011;36(5):614–624. doi: 10.1111/j.1365-2710.2010.01206.x. [DOI] [PubMed] [Google Scholar]
  • 45.Kurzawski M, Dąbrowska J, Dziewanowski K, Domański L, Perużyńska M, Droździk M. CYP3A5 and CYP3A4, but not ABCB1 polymorphisms affect tacrolimus dose-adjusted trough concentrations in kidney transplant recipients. Pharmacogenomics. 2014;15(2):179–188. doi: 10.2217/pgs.13.199. [DOI] [PubMed] [Google Scholar]
  • 46.Bruckmueller H, Werk AN, Renders L, et al. Which genetic determinants should be considered for tacrolimus dose optimization in kidney transplantation? A combined analysis of genes affecting the CYP3A locus. Ther. Drug Monit. 2015;37(3):288–295. doi: 10.1097/FTD.0000000000000142. [DOI] [PubMed] [Google Scholar]
  • 47.Pallet N, Jannot AS, El Bahri M, et al. Kidney transplant recipients carrying the CYP3A4*22 allelic variant have reduced tacrolimus clearance and often reach supratherapeutic tacrolimus concentrations. Am. J. Transplant. 2015;15(3):800–805. doi: 10.1111/ajt.13059. [DOI] [PubMed] [Google Scholar]
  • 48.Aouam K, Kolsi A, Kerkeni E, et al. Influence of combined CYP3A4 and CYP3A5 single-nucleotide polymorphisms on tacrolimus exposure in kidney transplant recipients: a study according to the post-transplant phase. Pharmacogenomics. 2015;16(18):2045–2054. doi: 10.2217/pgs.15.138. [DOI] [PubMed] [Google Scholar]
  • 49.Shi WL, Tang HL, Zhai SD. Effects of the CYP3A4*1B genetic polymorphism on the pharmacokinetics of tacrolimus in adult renal transplant recipients: a meta-analysis. PLoS ONE. 2015;10(6):e0127995. doi: 10.1371/journal.pone.0127995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Yousef AM, Qosa H, Bulatova N, et al. Effects of genetic polymorphism in CYP3A4 and CYP3A5 genes on tacrolimus dose among kidney transplant recipients. Iran J. Kidney Dis. 2016;10(3):156–163. [PubMed] [Google Scholar]
  • 51.de Jonge H, de Loor H, Verbeke K, Vanrenterghem Y, Kuypers DR. In vivo CYP3A4 activity, CYP3A5 genotype, and hematocrit predict tacrolimus dose requirements and clearance in renal transplant patients. Clin. Pharmacol. Ther. 2012;92(3):366–375. doi: 10.1038/clpt.2012.109. [DOI] [PubMed] [Google Scholar]
  • 52.Wang Z, Wu S, Chen D, et al. Influence of TLR4 rs1927907 locus polymorphisms on tacrolimus pharmacokinetics in the early stage after liver transplantation. Eur. J. Clin. Pharmacol. 2014;70(8):925–931. doi: 10.1007/s00228-014-1673-2. [DOI] [PubMed] [Google Scholar]
  • 53.Bosó V, Herrero MJ, Bea S, et al. Increased hospital stay and allograft dysfunction in renal transplant recipients with CYP2C19 AA variant in SNP rs4244285. Drug Metab. Dispos. 2013;41(2):480–487. doi: 10.1124/dmd.112.047977. [DOI] [PubMed] [Google Scholar]
  • 54.Damon C, Luck M, Toullec L, et al. predictive modeling of tacrolimus dose requirement based on high-throughput genetic screening. Am. J. Transplant. 2016;17(4):1008–1019. doi: 10.1111/ajt.14040. [DOI] [PubMed] [Google Scholar]
  • 55.Boivin AA, Cardinal H, Barama A, et al. Influence of SLCO1B3 genetic variations on tacrolimus pharmacokinetics in renal transplant recipients. Drug Metab. Pharmacokinet. 2013;28(3):274–277. doi: 10.2133/dmpk.dmpk-12-sh-093. [DOI] [PubMed] [Google Scholar]
  • 56.Jacobson PA, Oetting WS, Brearley AM, et al. Novel polymorphisms associated with tacrolimus trough concentrations: results from a multicenter kidney transplant consortium. Transplantation. 2011;91(3):300–308. doi: 10.1097/TP.0b013e318200e991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Oetting WS, Jacobson PA, Israni AK. Validation is critical for GWAS-based associations. Am. J. Transplant. 2017;17(2):318–319. doi: 10.1111/ajt.14051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Oetting WS, Wu B, Schladt DP, et al. Genome wide association study identifies the common variants in CYP3A4 and CYP3A5 responsible for variation in tacrolimus trough concentration in Caucasian kidney transplant recipients. Pharmacogen. J. 2017 doi: 10.1038/tpj.2017.49. Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Israni A, Leduc R, Holmes J, et al. Single-nucleotide polymorphisms, acute rejection, and severity of tubulitis in kidney transplantation, accounting for center-to-center variation. Transplantation. 2010;90(12):1401–1408. doi: 10.1097/TP.0b013e3182000085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Pulk RA, Schladt DS, Oetting WS, et al. Multigene predictors of tacrolimus exposure in kidney transplant recipients. Pharmacogenomics. 2015;16(8):841–854. doi: 10.2217/pgs.15.42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Li YR, van Setten J, Verma SS, et al. Concept and design of a genome-wide association genotyping array tailored for transplantation-specific studies. Genome Med. 2015;7:90. doi: 10.1186/s13073-015-0211-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Johnson AD, Handsaker RE, Pulit S, Nizzari MM, O'Donnell CJ, de Bakker PI. W. SNAP: a web-based tool for identification and annotation of proxy SNPs using HapMap. Bioinformatics. 2008;24(24):2938–2939. doi: 10.1093/bioinformatics/btn564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Hewett M, Oliver DE, Rubin DL, et al. PharmGKB: the pharmacogenetics knowledge base. Nucleic Acids Res. 2002;30(1):163–165. doi: 10.1093/nar/30.1.163. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Pharmacogenomics are provided here courtesy of Taylor & Francis

RESOURCES