A Bioinformatics Workflow for VUS Reclassification Using Multiple Tools

Satkın N. B., Morina E.

European Biotechnology Congress, Tirane, Arnavutluk, 11 - 13 Eylül 2025, cilt.9, sa.1, ss.23-24, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Cilt numarası: 9
  • Doi Numarası: 10.2478/ebtj-2025-0022
  • Basıldığı Şehir: Tirane
  • Basıldığı Ülke: Arnavutluk
  • Sayfa Sayıları: ss.23-24
  • Trakya Üniversitesi Adresli: Evet

Özet

Variants of uncertain significance (VUS) pose a challenge in interpreting medical genetics. In silico tools—computational methods—help predict whether a variant could be disease-causing. This approach does not require laboratory experiments and is particularly helpful in rare diseases where studies are limited. By combining different bioinformatics tools, we can gain a deeper understanding of how a variant affects a gene or protein and make more accurate clinical predictions. Methods and Results: We used a bioinformatics workflow to analyze 76 VUS found in a gene. As a rare disease example, we demonstrated how in silico tools can support variant classification. After filtering with REVEL and AlphaMissense, we selected 28 variants for further analysis. We utilized tools such as PolyPhen-2 and SIFT for functional effect prediction, ConSurf for conservation analysis, and gnomAD for population frequency data. We also applied SpliceAI for splicing prediction, DynaMut2 and HOPE for protein stability, and COACH-D to investigate binding sites. This provided us a clearer idea of the possible effects of each variant. Conclusion: Our study demonstrates that combining multiple in silico tools is effective for classifying VUS. This method supports better clinical decisions and shows the power of bioinformatics in modern genetics. Keywords: Pathogenicity prediction, protein dynamics, precision medicine, variant classification