Abstract
Dear Editor:
First we would like to thank Drs. Beuy Joob and Viroj Wiwanitkit for sharing ideas and suggestions in their letter (5) regarding our recent article, “AA IDO1 Variant Genotype (G2431A, rs3739319) Is Associated with Severe Dengue Risk Development in a DEN-3 Brazilian Cohort” (2) and also want to add some information hereunder.
1. We would like to share ideas on this report. First, the variant G to A of IDO1 is a genetic change that can result in molecular change. The effect on the final phenotypic expression might be expected similar to the observed phenomenon in reported clinical interrelationship of other A–G altering genetic variant (4).
Answer: The SNP rs3739319 (G2431A) is an intronic variant that is involved with the enzyme transcriptional regulation that suggests a functional relevance (6). Thus, a shift from G to A maybe results in changes in the protein expression pattern that may compromise NAD+ coenzyme biosynthesis. Consequently, this may influence the functioning of several important metabolic pathways involved in inflammation, oxidative stress, and cell death (9).
2. Nevertheless, in the present cohort by Azevedo et al. (2), based on Hardy Weinberg genetic equilibrium (7,8), there is no normal distribution and the Yates correction was used (2). Nevertheless, the derived odds ratio should be further adjusted. Confounding adjusted odds ratio should be calculated (10). At least the adjustment for confounding factors is needed.
Answer: The Yates test was used to verify the degree of continuity of significance for equal proportions among dengue groups and significance was observed among the studied groups (table 3 of the article) (2). In fact, we do not develop the multivariated logistic regression analysis in this article because we performed logistic regression and other statistical methods in our previous article using this same cohort and features, such as age, sex, and immune history. (1).
The interaction of genetic polymorphisms are real confounding factors. Recently our group published an article on the prognosis of severe dengue using machine learning, where it is highlighted that the genetic context may be the “key element” for genetically influenced diseases (3).
3. The important confounding factors include demographic background (age, sex, ethnic, etc.) as well as other genetic factors that might affect dengue severity (such as MICB, TNF, CD209, FcγRIIA, TPSAB1, CLEC5A, IL10, and PLCE1 gene variants (12)). Finally, the severity of dengue is not totally based on genetic underlying factors. For example, shock can be controlled if there is a good early fluid replacement therapy for dengue patients (11). Good clinical management can result in good clinical outcome of dengue care (11).
Answer: We agree with your idea about the realization of stratification. In previous study (1) we used this method in the same cohort (∼100 samples), and we could observe that there is a growing risk for severe dengue as the age advances. As commented before, we believe that there are many confounding factors in this genetically influenced disease. Some of them (genomic features) were discussed in our previous article using genome data and machine learning techniques to dengue prognosis (3).
For ancestry and ethnicity we do not intend to develop this analysis.