Computational genomics in rare fibrinogen deficiencies leading to hemorrhagic psychosis

Dear Editor, Fibrinogen is a liver-derived plasma protein consisting of three polypeptide chains (Aα, Bβ, and γ) encoded by the FGA, FGB, and FGG genes and pathogenic mutations in these genes can cause fibrinogen deficiency. More than 200 pathogenic variants have been reported in the FGB gene, many affecting the conserved C-terminal region of the β-chain. Genetic screening revealed a novel nonsense mutation in exon 8 of the FGB gene (c.1299G>A; p.Trp433*) that was present across four generations. This mutation introduces a premature stop codon likely resulting in β-chain truncation or nonsense-mediated mRNA decay[1]. Clinically, severe fibrinogen deficiency can make patients prone to intracerebral hemorrhage, a complication that can rarely present with neuropsychiatric manifestations such as hemorrhagic psychosis[2]. The Genome Aggregation Database (gnomAD) is a key reference for genetic variant interpretation, providing large-scale population allele frequency data. Using gnomAD v4.1.0 data from over 800 000 individuals, recent analyses suggest that both recessive and dominant fibrinogen deficiencies are more common than previously reported. The use of updated gnomAD datasets allows clinicians to better recognize underdiagnosed fibrinogen deficiencies and supports more informed interpretation of genetic findings. Next-generation sequencing approaches, including whole-exome sequencing and whole-genome sequencing, allow comprehensive detection of both primary pathogenic variants and additional genetic modifiers[3]. Advances in computational genomics further allow integration of sequencing data with routine clinical biomarkers, such as fibrinogen levels, to build patient-specific models of coagulation[4]. Previous studies have used sequencing of FGA, FGB, and FGG to identify distinct pathogenic variants, including recurrent mutations in FGA and FGG, which help classify the subtypes of congenital fibrinogen disorders and associate specific genetic changes with clinical features[5]. Moreover, sequencing of the FGB gene, supported by in silico analyses and genomic databases, has enabled identification of pathogenic variants such as c.1299G>A across multiple generations[1]. The wider use of next-generation sequencing coupled with comprehensive genomic data has greatly improved research on congenital fibrinogen deficiencies, allowing identification of additional variants beyond primary pathogenic mutations. Analyses using large datasets such as gnomAD v4.1.0 have refined prevalence estimates, capturing rare variants and showing higher frequencies than earlier believed[3]. Estimating disease prevalence from gnomAD v4 data is challenging because the presence of a genotype does not always indicate disease, and the analysis focuses only on coding regions and splice sites, overlooking noncoding regulatory elements that may contribute to fibrinogen deficiencies[3]. In addition, current computational approaches often fail to fully incorporate such personalized genomic and biomarker information, thereby restricting their predictive capability. Improving these tools by combining genetic information with patient-specific clinical data is essential for better application of precision medicine in bleeding disorders. Computational coagulation models provide valuable insight into clotting dynamics and support clinical decision-making. Expanding such models to incorporate fibrinogen levels and genomic data would improve prediction of bleeding risk and optimize management in congenital fibrinogen deficiencies[4]. Computational genomics provides key insights into rare fibrinogen deficiencies and associated complications like hemorrhagic psychosis. The integration of genomic information with patient-specific biomarkers improves prediction of bleeding risk and management strategies. The understanding of rare variants and the incorporation of noncoding regions into fibrinogen deficit studies are areas that require further investigation for advancing precision medicine in congenital fibrinogen disorders. This letter to the editor adheres to the Transparency in the Reporting of Artificial Intelligence in Research (TITAN) guideline[6].