LAUSR

Abstract Background Neurofibromatosis type 1 (NF-1) is a tumor predisposition syndrome resulting from mutations in NF1 antioncogene. Differential diagnosis with genetic testing is essential in NF-1 diagnostics due to its clinical variability. Identifying of pathogenic mutations is challenging because of the gene size and structure, especially for low-level somatic mosaicism. Notably, patients with mosaic NF-1 demonstrate no less severe phenotype than those with classic form. Nowadays, NGS and Sanger sequencing are recognized diagnostics methods and they both are not sensitive enough to mosaic genotypes with small fractions of alternative alleles. We believe that this can be resolved with a special approach to sequencing data analysis. Methods In order to re-evaluate previously obtained NGS (Ion AmpliSeq technology) results for 275 probands with clinical diagnosis “NF-1” or “NF, unspecified” for whom germline mutations in NF1 and NF2 genes were not identified, we developed an improved data analysis pipeline to search for somatic mutations, which includes: (1) programmatic pool-based division of NGS reads, (2) elimination of out-of-design aligned reads, (3) exclusion of systematic sequencing errors, (4) variant calling with low stringency parameters (alternative allele frequency, AF ≥ 0.05; read depth, DP ≥ 20). Newly detected mutations were verified using Sanger sequencing and heteroduplex analysis. Sanger results were analysed using our in-house SeqBase software, highly sensitive to mosaic variants. Results Application of our NGS data reanalysis algorithm allowed us to detect 12 cases (4.3%) of somatic mosaic mutations among 275 probands otherwise lacking molecular verification of neurofibromatosis diagnosis. The majority of the identified mutations are nonsense or frameshift, with AF ranging from 0.051 to 0.296. Mutations were verified with alternative methods of molecular diagnostics. Conclusions Improved sequencing data analysis allows to detect NF-1 cases with mosaic genotype in cases unresolved by conventional analysis pipeline. Yet, success rate is strongly dependent upon the fraction of the mutant allele, thus alternative quantitative methods are required for exhaustive NF-1 molecular diagnosis. Legal entity responsible for the study Federal State Budgetary Institution "Research Centre for Medical Genetics". Funding The research was carried out within the state assignment of Ministry of Science and Higher Education of the Russian Federation. Disclosure All authors have declared no conflicts of interest.