Brian Rhees

Localization of a susceptibility gene for type 2 diabetes to chromosome 5q34-q35.2.

We report a genomewide linkage study of type 2 diabetes (T2D [MIM 125853]) in the Icelandic population. A list of type 2 diabetics was cross-matched with a computerized genealogical database clustering 763 type 2 diabetics into 227 families. The diabetic patients and their relatives were genotyped with 906 microsatellite markers. A nonparametric multipoint linkage analysis yielded linkage to 5q34-q35.2 (LOD = 2.90, P=1.29 x 10(-4)) in all diabetics. Since obesity, here defined as body mass index (BMI) > or =30 kg/m(2), is a key risk factor for the development of T2D, we studied the data either independently of BMI or by stratifying the patient group as obese (BMI > or =30) or nonobese (BMI <30). A nonparametric multipoint linkage analysis yielded linkage to 5q34-q35.2 (LOD = 3.64, P=2.12 x (10)-5) in the nonobese diabetics. No linkage was observed in this region for the obese diabetics. Linkage analysis conditioning on maternal transmission to the nonobese diabetics resulted in a LOD score of 3.48 (P=3.12 x 10(-5)) in the same region, whereas conditioning on paternal transmission led to a substantial drop in the LOD score. Finally, we observed potential interactions between the 5q locus and two T2D susceptibility loci, previously mapped in other populations.

The effect of insulin on expression of genes and biochemical pathways in human skeletal muscle.

To study the insulin effects on gene expression in skeletal muscle, muscle biopsies were obtained from 20 insulin sensitive individuals before and after euglycemic hyperinsulinemic clamps. Using microarray analysis, we identified 779 insulin-responsive genes. Particularly noteworthy were effects on 70 transcription factors, and an extensive influence on genes involved in both protein synthesis and degradation. The genetic program in skeletal muscle also included effects on signal transduction, vesicular traffic and cytoskeletal function, and fuel metabolic pathways. Unexpected observations were the pervasive effects of insulin on genes involved in interacting pathways for polyamine and S-adenoslymethionine metabolism and genes involved in muscle development. We further confirmed that four insulin-responsive genes, RRAD, IGFBP5, INSIG1, and NGFI-B (NR4A1), were significantly up-regulated by insulin in cultured L6 skeletal muscle cells. Interestingly, insulin caused an accumulation of NGFI-B (NR4A1) protein in the nucleus where it functions as a transcription factor, without translocation to the cytoplasm to promote apoptosis. The role of NGFI-B (NR4A1) as a new potential mediator of insulin action highlights the need for greater understanding of nuclear transcription factors in insulin action.

Association of Phosphodiesterase 4D Polymorphisms With Ischemic Stroke in a US Population Stratified by Hypertension Status

Background and Purpose— Phosphodiesterase 4D (PDE4D) underlies the STRK1 linkage peak for stroke on chromosome 5q12 identified in Iceland. We tested association of 13 single-nucleotide polymorphisms (SNPs) and 1 microsatellite in a nested case-control sample of elderly white women (>65 years of age) from the Study of Osteoporotic Fractures (SOF) in the United States. Methods— The genotypes of 248 women who experienced an incident ischemic stroke during an average of 5.4 years of follow-up were compared with 560 controls. Results— Marginal associations with stroke (P<0.10) were found for 3 polymorphisms. Stratification of the population by hypertension markedly strengthened the association. SNPs 9 (hazard ratio [HR], 0.48; 95% CI, 0.26 to 0.91), 42 (HR, 1.73; 95% CI, 1.10 to 2.70), 219 (HR, 1.73; 95% CI, 1.13 to 2.64), and 220 (HR, 1.56; 95% CI, 1.05 to 2.32) showed significant association with stroke (P<0.05) under a dominant model in subjects without hypertension at baseline, and SNP 175 was significantly associated with stroke under an additive model (HR, 0.76; 95% CI, 0.59 to 0.98) in subjects with hypertension. Furthermore, the microsatellite AC008818-1 showed association with stroke only in the nonhypertensive subjects. Based on results in Iceland, specific haplotypes were tested in SOF, and stratification by hypertension also affected these association results. Conclusion— These data are consistent with an association of the PDE4D gene with stroke in a non-Icelandic sample and suggest an effect of hypertension status.

Genomics approaches to drug discovery

New approaches to drug discovery have come about in recent years as a result of important advances in genomics and bioinformatics. The availability of genome‐scale sequence data, the development of new tools for high‐throughput gene expression monitoring, and improvements in the ability to analyze large data sets have revolutionized the field. In this article, we discuss three applications of genomics data in the drug discovery process: target discovery, prodrug strategies, and vaccine development. J. Cell. Biochem. Suppl. 37: 110–119, 2001.

Reply to Kraft

To the Editor:Our study (Bugawan et al. 2003xAssociation and interaction of the IL4R, IL4, and IL13 loci with type 1 diabetes among Filipinos. Bugawan, TL, Mirel, DB, Valdes, AM, Panelo, A, Pozzilli, P, and Erlich, HA. Am J Hum Genet. 2003; 72: 1505–1514Abstract | Full Text | Full Text PDF | PubMed | Scopus (48)See all References2003) reported a negative association of a specific IL4-524 haplotype with type 1 diabetes (T1D), consistent with a previous report (Mirel et al. 2002xAssociation of IL4R haplotypes with type 1 diabetes. Mirel, DB, Valdes, AM, Lazzeroni, LC, Reynolds, RL, Erlich, HA, and Noble, JA. Diabetes. 2002; 51: 3336–3341Crossref | PubMedSee all References2002), and presented evidence for a genetic interaction between IL4-524 and IL4R SNPs. To test the latter, we computed relevant P values by permuting multilocus genotypes separately in case and control groups.The criticism raised by Kraft (2004xMultiple comparisons in studies of gene × gene, gene × environment interaction. Kraft, P. Am J Hum Genet. 2004; 74: 582–584Abstract | Full Text | Full Text PDF | PubMed | Scopus (9)See all References2004 [in this issue]) is not directed at our implementation of permutation testing, per se, but at permutation testing in general. His argument is that permutation testing does not properly account for multiple comparisons, resulting in an increase in false claims of significance, or type I familywise error (FWE). In the place of permutation testing, Kraft advocates the use of the Simes method—an elaboration of the classic Bonferroni procedure. In response, we wish to show that permutation testing can be used to obtain a desired false-positive error rate (as, indeed, can be demonstrated using Kraft’s example) and, moreover, that such an approach has the added advantage of providing additional protection against false claims of nonsignificance, or type II error.It should be noted that permutation methods are well established as a robust approach for obtaining overall significance levels while minimizing type II error (e.g., Good 1994xGood, P. CrossrefSee all References1994; Doerge and Churchill 1996xPermutation tests for multiple loci affecting a quantitative character. Doerge, RW and Churchill, GA. Genetics. 1996; 142: 285–294PubMedSee all References1996; Lynch and Walsh 1998xLynch, M and Walsh, B. : 441–442See all References1998), that such methods are extensible to multiple-testing scenarios (Westfall and Young 1993xWestfall, PH and Young, SS. See all References1993), and that examples of their application to human genetics are not uncommon (e.g., Lewis et al. 2003xGenome scan meta-analysis of schizophrenia and bipolar disorder, part II: schizophrenia. Lewis, CM, Levinson, DF, Wise, LH, DeLisi, LE, Straub, RE, Hovatta, I, Williams, NM et al. Am J Hum Genet. 2003; 73: 34–48Abstract | Full Text | Full Text PDF | PubMed | Scopus (807)See all References2003). However, as with any statistical method, the validity is dependent on correct application. Kraft provides an analysis of the permutation testing by discussing the distribution of two P values obtained from hypothetically permuted distributions (i.e., independent and uniformly distributed under the null hypothesis). The joint cumulative distribution function (CDF) for these two P values is given as F(P(1),P(2))=P(1)(2P(2)−P(1)), where P(1) and P(2) are, respectively, the first- and second-ordered P values. As such, Kraft notes that the Pr(P<.05) for this joint distribution is ∼0.1, indicating that we would expect to see the smaller P value, or P(1)<.05, about 10% of the time. Kraft’s argument, therefore, is that for independent tests, use of a critical value of .05 leads to a type I error rate of 10%.In fact, the proper approach for permutation testing—adjusted or unadjusted for multiple comparisons—is to find the critical value corresponding to the desired type I error rate. Specifically, if we consider the simulations presented by Kraft as equivalent to the result of a permutation test, we would seek the value of x in the permuted distribution for which Pr(P<x) is actually ≤α and would use that value, not the .05 value as Kraft appears to suggest. For P(1), this critical value would be .0253, as can be shown either by simulation or by solving Kraft’s joint CDF for α=0.05, given P(2)=1 (in effect, solving the marginal CDF for P(1)). It is interesting to note that the first P value that Kraft gives (.10) corresponds to the Sidak multiple comparison–adjusted P value for observed α=0.05 and k=2 tests, whereas the value we give corresponds to the Sidak-adjusted threshold (1−[1−α]1/k). As such, this example nicely illustrates that permutation testing, for two independent tests, yields familiar and contextually appropriate results.It should also be noted that multiple-testing methods that rely on raw Bonferroni-type inequalities fail to incorporate correlation structures between tests. Therefore, although such methods (e.g., Simes 1986xAn improved Bonferroni procedure for multiple tests of significance. Simes, RJ. Biometrika. 1986; 73: 751–754Crossref | Scopus (918)See all References1986; Hochberg 1988xA sharper Bonferroni procedure for multiple tests of significance. Hochberg, Y. Biometrika. 1988; 75: 800–802Crossref | Scopus (2414)See all References1988; Rom 1990xA sequentially rejective test procedure based on a modified Bonferroni inequality. Rom, DM. Biometrika. 1990; 77: 663–665Crossref | Scopus (218)See all References1990) provide control of FWE, they nevertheless are expected to be less powerful than methods that account for such dependencies. Indeed, these methods may be made more precise through resampling-based approaches (Westfall and Young 1993xWestfall, PH and Young, SS. See all References1993). In particular, the data from which the tests in table 7 (Bugawan et al. 2003xAssociation and interaction of the IL4R, IL4, and IL13 loci with type 1 diabetes among Filipinos. Bugawan, TL, Mirel, DB, Valdes, AM, Panelo, A, Pozzilli, P, and Erlich, HA. Am J Hum Genet. 2003; 72: 1505–1514Abstract | Full Text | Full Text PDF | PubMed | Scopus (48)See all References2003) were derived are strongly correlated, and, therefore, tests that assume independence are not expected to be the most powerful. Moreover, Kraft fails to take into account the nonindependence of genotype distributions between chromosome 5 and chromosome 16 SNPs presented in table 6 (Bugawan et al. 2003xAssociation and interaction of the IL4R, IL4, and IL13 loci with type 1 diabetes among Filipinos. Bugawan, TL, Mirel, DB, Valdes, AM, Panelo, A, Pozzilli, P, and Erlich, HA. Am J Hum Genet. 2003; 72: 1505–1514Abstract | Full Text | Full Text PDF | PubMed | Scopus (48)See all References2003). Applying the Simes correction suggested by the author for 10 comparisons (two sets: patients and controls, and five SNPs), the independence between IL4-524 and IL4R patient genotypes would be rejected with P<.01, supporting our conclusion of an interaction between chromosome 5 and chromosome 16 in T1D susceptibility.In conclusion, what is needed, from a methodological perspective, are statistical procedures that adequately protect against false claims of significance while simultaneously addressing the correlated nature of multiple testing. The various methods discussed by Kraft address the former but do not address the latter. Having said this, whatever the statistical approach, the strongest test of the significance of any reported genetic interaction lies neither in initial-discovery P values nor in biologic plausibility—which we believe is high in this case—but in the ability to reproduce observations in independent cohorts.