Commentary: Long Non-Coding RNA Gene Polymorphisms and Their Expression Levels in Patients With Rheumatoid Arthritis

A Commentary on
Long Non-Coding RNAs Genes Polymorphisms and Their Expression Levels in Patients With Rheumatoid Arthritis

By Zhang T-P, Zhu B-Q, Tao S-S, Fan Y-G, Li X-M, Pan H-F, et al. (2019). 10:2529. doi: 10.3389/fimmu.2019.02529

Introduction

We read the paper by Zhang et al. (1) with interest. The authors report on a study that evaluated the association of four long noncoding RNA (lncRNA) (ANRIL, lnc-DC, MALAT1, ZFAS1) gene single-nucleotide polymorphisms (SNPs) with susceptibility to rheumatoid arthritis (RA) patients, as well as their expression levels. They concluded that ANRIL, lnc-DC, MALAT1, and ZFAS1 gene SNPs were not associated with RA susceptibility after false discovery rate (FDR) correction, while altered ANRIL, lnc-DC, MALAT1, and ZFAS1 levels in RA patients suggested that these lncRNAs might play a role in RA.

After carefully reading, we identified some mistakes in the odds ratio (OR) calculations in Table 1 “Genotypes and alleles frequencies of lncRNAs genes polymorphisms in RA patients and normal controls” [sic]. Likewise, there are inconsistencies in the genetic models.

TABLE 1

Table 1 Correction: Genotypes and allele frequencies of lncRNA gene polymorphisms in RA patients and normal controls.

In general, we find that the OR values were calculated and interpreted in an inappropriate way. This is very noticeable when analyzing the frequency of genotypes in cases and controls. For example, for rs1412830, the authors reported an OR = 0.214 (0.060–0.761), p = 0.017, for the TT genotype. That OR value suggests that this genotype is a protective or lower risk factor. However, the TT genotype has a higher prevalence in patients than that in controls (1.97% vs. 0.42%, respectively); therefore, the OR value should be >1.

Regarding genetic models, the authors maintain the same mistake in the OR values. Taking rs1412830 as an example, we once again observed that there is a higher frequency of the TT genotype in patients than that in controls (1.97% vs. 0.42%, respectively). However, the authors reported an OR value = 0.211 (0.059–0.750), p = 0.016, but they describe it as a risk factor. We correctly calculate the OR values for your consideration (Table 1).

We recommend that the authors (1) recalculate these data appropriately (2) in order to be able to rediscuss all their results. Also, we suggest that they corroborate the OR values in Table 3 (1), which we could not analyze due to lack of data.

Author Contributions

JH-B wrote this commentary and analyzed the data. CB-H and JM-V performed the analysis. All authors contributed to the article and approved the submitted version.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Zhang T-P, Zhu B-Q, Tao S-S, Fan Y-G, Li X-M, Pan H-F, et al. Long Non-Coding RNAs Genes Polymorphisms and Their Expression Levels in Patients With Rheumatoid Arthritis. Front Immunol (2019) 10:2529. doi: 10.3389/fimmu.2019.02529

PubMed Abstract | CrossRef Full Text | Google Scholar

2. Sato Y, Suganami H, Hamada C, Yoshimura I, Sakamoto H, Yoshida T, et al. The Confidence Interval of Allelic Odds Ratios Under the Hardy–Weinberg Disequilibrium. J Hum Genet (2006) 51:772–80. doi: 10.1007/s10038-006-0020-6

PubMed Abstract | CrossRef Full Text | Google Scholar