R202Q dominates MEFV variant spectrum in Eastern Black Sea suspected FMF cohort

A single-centre study from the Eastern Black Sea region of Türkiye has mapped the spectrum of variants in the MEFV gene among 611 consecutive individuals whose genetic test reports were analysed between January 2022 and October 2025 after referral with a clinical suspicion of familial Mediterranean fever. The retrospective analysis of next-generation sequencing reports, published in BMC Genomic Data, found that at least one MEFV variant was present in 379 of the 611 individuals, or 62.0 percent, while no reportable variant was identified in the remaining 232 individuals, or 38.0 percent. Across the variant-positive group, 591 mutant alleles were tabulated, spanning 22 distinct missense variants, and the resulting frequency table offers one of the most detailed molecular portraits yet published for this part of the Turkish Black Sea coast.

The headline finding is a striking imbalance at the top of the variant spectrum. R202Q, formally c.605G>A, p.Arg202Gln, an exon 2 substitution whose disease-causing status has long been debated, was the most frequently detected allele, accounting for 44.0 percent of all detected mutant alleles, with 260 of the 591 alleles, and appearing on 21.3 percent of all 1,222 tested chromosomes. M694V, the canonical exon 10 founder mutation c.2080A>G, ranked second at 132 alleles, or 22.3 percent of detected alleles and 10.8 percent of tested chromosomes. E148Q followed with 55 alleles (9.3 percent of detected alleles; 4.5 percent of tested chromosomes), then M680I with 43 alleles (7.3 percent; 3.5 percent) and V726A with 20 alleles (3.4 percent; 1.6 percent). Among rarer detections, P369S and R408Q each appeared in 17 alleles, typically co-occurring on the same chromosome, A744S in six alleles, and p.Ala89Thr and p.Leu509Pro in one allele each.

The authors are careful to separate spectrum prominence from pathogenic burden, and this distinction drives much of the paper’s argument. When the detected alleles were stratified by category of pathogenic evidence, the high-penetrance exon 10 founder variants M694V, M680I, M694I and V726A together accounted for only 202 alleles, corresponding to 34.2 percent of the detected mutant-allele pool but just 16.5 percent of all tested chromosomes. E148Q, generally regarded as a low-penetrance disease-modifying variant, contributed 55 alleles, and R202Q, a variant of contested pathogenicity, contributed 260. The remaining uncommon variants together added 74 alleles, or 12.5 percent of the detected pool and 6.1 percent of tested chromosomes. Because many experts in the field regard R202Q as a benign polymorphism rather than a disease-causing variant, the study concludes that the dominance of this allele in the detected spectrum should not be equated with the regional clinical burden of familial Mediterranean fever or with the burden of pathogenic alleles.

This regional pattern is consistent with earlier Black Sea data. Multicentre studies have shown M694V to be the most frequently detected mutation in Turkish familial Mediterranean fever patients overall, but regional cohorts from Central Black Sea provinces have previously reported unexpectedly high prevalence of R202Q as the predominant detected allele. In the Mid-Black Sea region, Akbaş and colleagues identified R202Q as the most frequent variant at 27.9 percent, and a genotype-level study from Amasya by Çapraz and Düz reported an R202Q frequency of 43.5 percent. The authors interpret the Eastern Black Sea predominance of R202Q as a regional detection and spectrum pattern, possibly reflecting founder effects and reduced genetic admixture relative to interior Anatolian populations, rather than as evidence of a regionally dominant pathogenic mutation. They explicitly make no recommendation that R202Q be used as a diagnostic criterion.

The zygosity and genotype architecture of the cohort adds another layer of complexity. Of the 591 tabulated alleles, 561, or 94.9 percent, derived from heterozygous reports and 30, or 5.1 percent, from homozygous reports, corresponding to 15 homozygous detections. Homozygosity was observed for R202Q in nine detections, M694V in four and M680I in two, while no homozygous detection was recorded for E148Q despite its 55 alleles or for V726A despite its 20. Compound heterozygous or complex multi-variant genotypes were present in 157 individuals, or 25.7 percent of the total cohort and 41.4 percent of the variant-positive group, while 216 individuals carried a single heterozygous variant and six carried a single homozygous variant.

A critical caveat shadows all genotype interpretation: cis/trans phase was not determined, because parental segregation testing was not available. Two well-recognised cis haplotypes are expected within these data, since M694V is frequently in linkage disequilibrium with R202Q on the same haplotype and P369S co-occurs in cis with R408Q. An apparent two-variant genotype may therefore reside on a single mutant chromosome. On an unphased basis, 126 of the 157 compound or complex genotypes, or 80.3 percent, contained at least one high-penetrance exon 10 allele, and 24 of these carried two or more exon 10 alleles, a strongly biallelic high-penetrance configuration even without phasing. Yet the single most frequent configuration was M694V plus R202Q, seen in 77 individuals, precisely the pairing for which the recognised cis haplotype means many may represent a single pathogenic chromosome rather than a biallelic pathogenic genotype. The remaining 31 genotypes, or 19.7 percent, were composed exclusively of exon 2 and uncommon variants, most frequently E148Q plus P369S plus R408Q and E148Q plus R202Q, representing a lower-penetrance, non-confirmatory configuration. Separately, three individuals were homozygous for a high-penetrance exon 10 variant as their sole finding.

The study also applied an exploratory computational layer to variants outside the recognised common familial Mediterranean fever alleles. Fourteen uncommon missense variants were submitted to a seven-tier classification framework published by Alay in 2025, which integrates 42 in silico predictors catalogued in dbNSFP and Ensembl through an ensemble machine-learning model trained on curated MEFV variants. The framework offers an adaptive classifier, which collapses first- and second-tier uncertain calls into likely-benign or likely-pathogenic categories, and a more conservative rigid classifier, which collapses only the strongest calls. Ten of the 14 variants returned an output; under the adaptive classifier, seven were flagged toward likely-pathogenic, namely E148V, E167D, S339F, F479L, K695R, A744S and R761H, and three toward likely-benign, namely T267I, G304R and I591T. Under the rigid classifier, all seven likely-pathogenic flags were retained and two of the three likely-benign flags were preserved, with only T267I shifting to indeterminate. Four variants, including the novel p.Ser179Gly, p.A287T, p.Ala89Thr and p.Leu509Pro, had no exact row in the deployed application and could not be ranked.

The authors stress repeatedly that these computational flags are exploratory and do not constitute clinical reclassification. Under the 2015 ACMG/AMP guidelines, and in the absence of functional, segregation and phenotype data, all of the uncommon variants remain variants of uncertain significance. Cross-referencing against the expert INFEVERS database showed broad agreement for several variants, with F479L and R761H classified as pathogenic and G304R and I591T as benign or likely benign, but also clear discordance for A744S and S339F, which the classifier flagged as likely-pathogenic yet INFEVERS lists as likely benign. The paper points to newly available high-throughput functional data from a method called SpeckSeq, which functionally stratified a large panel of MEFV variants and identified 49 gain-of-function substitutions, as the kind of orthogonal evidence that could corroborate or refute the in silico flags before any clinical use.

One detection is reported with particular caution. A previously unreported missense substitution, p.Ser179Gly, was found in a single individual in apparent compound configuration with R202Q. A literature search did not identify a prior peer-reviewed clinical report of this specific substitution, although the neighbouring p.Ser179Asn is registered in ClinVar with conflicting evidence. Because orthogonal Sanger confirmation and quantitative raw-read quality metrics were not available for this retrospective report, and because a single previously unreported sequencing call cannot be distinguished from a technical artefact without confirmatory testing, the variant is reported strictly as a candidate variant of uncertain significance with no pathogenicity inference drawn from it.

The limitations of the work frame how its findings should be read. No standardised diagnostic or classification criteria, neither Tel-Hashomer nor Eurofever/PRINTO, were applied, and attack characteristics, colchicine response, family history and amyloidosis status were not collected, so the cohort represents individuals tested for suspected familial Mediterranean fever rather than clinically confirmed patients, and the analysis is framed as a molecular variant-spectrum study rather than a clinical cohort study. The frequencies describe a referral, test-indication population and are not estimates of disease prevalence or of pathogenic-allele burden in the general regional population. The targeted sequencing method, covering coding exons and ten base pairs of flanking sequence with a 20-read minimum, was not designed to detect deep intronic variants, large copy-number alterations or structural variants. Within those constraints, the authors describe this as the largest MEFV variant dataset reported from this Eastern Black Sea province and the first to characterise its variant spectrum alongside an exploratory gene-specific in silico stratification, with the full anonymised dataset openly available.

Subject of Research: MEFV gene variant spectrum and exploratory in silico stratification of uncommon variants in individuals tested for suspected familial Mediterranean fever in the Eastern Black Sea region of Türkiye

Article Title: MEFV variant spectrum and exploratory in silico stratification of uncommon variants in 611 individuals tested for suspected familial Mediterranean fever in the Eastern Black Sea region of Türkiye

Article References: Doğan, Ç. (2026). MEFV variant spectrum and exploratory in silico stratification of uncommon variants in 611 individuals tested for suspected familial Mediterranean fever in the Eastern Black Sea region of Türkiye. BMC Genomic Data, 27(1), Article 79. https://doi.org/10.1186/s12863-026-01481-z

Image Credits: AI Generated

DOI: 10.1186/s12863-026-01481-z

Keywords: MEFV, familial Mediterranean fever, R202Q, M694V, genetic variants, Türkiye, Black Sea region, next-generation sequencing, in silico classification, variants of uncertain significance, pyrin, ACMG/AMP