Purpose: According to the American College of Medical Genetics (ACMG) classification, variants of uncertain significance (VUS) are gene variations whose impact on the disease risk is not yet known. VUS, therefore, represent an unmet need for genetic counselling. Aim of the study is the use the AlphaFold artificial intelligence algorithm to predict the impact of novel mutations of the IGFALS gene, detected in a subject with short stature and initially classified as VUS according to the ACMG classification. Methods: A short-stature girl and her parents have been investigated. IGFALS mutations have been detected through clinical exome and confirmed by Sanger sequencing. The potential presence of co-occurring gene alterations was investigated in the proband by whole exome and CGH array. Structure of the ALS protein (encoded by the IGFALS gene) was evaluated through the AlphaFold artificial intelligence algorithm. Results: Two IGFALS variants were found in the proband: c.1349T > C (p.Leu450Pro) and c.1363_1365delCTC (p.Leu455del), both classified as VUS, according to ACMG. Parents’ analysis highlighted the in trans position of the two variants. AlphaFold showed that the mutated positions were found the concave side a horseshoe structure of the ALS protein, likely interfering with protein-protein interactions. According to a loss of function (LoF) effect of the two variants, reduced levels of the IGF1 and IGFBP-3 proteins, as well as a growth hormone (GH) excess were detected in the proband’s serum. Conclusions: By using the AlphaFold structure we were able to predict two IGFALS gene mutations initially classified as VUS, as potentially pathogenetic. Our proof-of-concept showed a potential application of AlphaFold as tool to a better inform VUS interpretation of genetic tests.
