Case Report: The Role of Molecular Analysis of the MUTYH Gene in Asymptomatic Individuals

DARVIC: Dihedral angle-reliant variant impact classifier for functional prediction of missense VUS

BACKGROUND

Of the large number of genetic variants identified, the functional impact for most of them remains unknown. Mutations in DNA damage repair genes such as MUTYH, which is involved in repairing A:8-oxoG mismatches caused by reactive oxygen species, can cause a higher risk of cancer. Mutations happening in other key genes such as TP53 also pose huge health threats and risk of cancer. The interpretation of genetic variants' functional impact is a forefront issue that needs to be addressed. Many different in silico methods based on different principles have been developed and applied in interpreting genetic variants. However, a current challenge is that many existing methods tend to overpredict the pathogenicity of benign variants. A new approach is needed to tackle this issue to improve genetic variant interpretation through the use of in silico methods.

METHODS

In this study, we developed another protein structural-based approach called Dihedral angle-reliant variant impact classifier (DARVIC) to predict the deleterious impact of the coding-changing missense variants. DARVIC uses Ramachandran's principle of protein stereochemistry as the theoretical foundation and uses molecular dynamics simulations coupled with a supervised machine learning algorithm XGBoost to determine the functional impact of missense variants on protein structural stability.

RESULTS

We characterized the features of dihedral angles in dynamic protein structures. We also tested the performance of DARVIC in MUTYH and TP53 missense variants and achieved satisfactory results in reflecting the functional impacts of the variants on protein structure. The method achieved a balanced accuracy of 84% in a functionally validated MUTYH dataset containing both benign and pathogenic missense variants, higher than other existing in silico methods. Along with that, DARVIC was able to predict 119 (47%) deleterious variants from a dataset of 254 MUTYH VUS. Further application of DARVIC to a functionally validated TP53 dataset had a balanced accuracy of 94%, topping other methods, demonstrating DARVIC's robustness.

CONCLUSION

DARVIC provides a valuable tool to predict the functional impacts of missense variants based on their effects on protein structural stability and motion. At its current state, DARVIC performed well in predicting the functional impact of the missense variants both in MUTYH and TP53. We expect its high potential to predict functional impact for the missense variants in other genes.

Cancer Predisposition Genes in Adolescents and Young Adults (AYAs): a Review Paper from the Italian AYA Working Group

PURPOSE OF REVIEW

The present narrative systematic review summarizes current knowledge on germline gene mutations predisposing to solid tumors in adolescents and young adults (AYAs).

RECENT FINDINGS

AYAs with cancer represent a particular group of patients with specific challenging characteristics and yet unmet needs. A significant percentage of AYA patients carry pathogenic or likely pathogenic variants (PV/LPVs) in cancer predisposition genes. Nevertheless, knowledge on spectrum, frequency, and clinical implications of germline variants in AYAs with solid tumors is limited. The identification of PV/LPV in AYA is especially critical given the need for appropriate communicative strategies, risk of second primary cancers, need for personalized long-term surveillance, potential reproductive implications, and cascade testing of at-risk family members. Moreover, these gene alterations may potentially provide novel biomarkers and therapeutic targets that are lacking in AYA patients. Among young adults with early-onset phenotypes of malignancies typically presenting at later ages, the increased prevalence of germline PV/LPVs supports a role for genetic counseling and testing irrespective of tumor type.

Actionable secondary findings in the 73 ACMG-recommended genes in 1559 Thai exomes

High-throughput DNA sequencing provides not only primary diagnosis but also makes available other genetic variants with potential health implications. the American College of Medical Genetics and Genomics (ACMG) has recommended a list of medically actionable genes since 2013 and very recently released an updated ACMG SF v3.0 list comprising 73 genes. Here, we analyzed exome data of 1559 unrelated Thai individuals to determine the frequency and spectrum of pathogenic (P) or likely pathogenic (LP) variants in the 73 genes. Based on the ACMG guidelines for the interpretation of sequence variants, 68 different P/LP variants in 26 genes associated with 18 diseases inherited in an autosomal-dominant manner of 186 individuals (11.9%; 186/1559) were identified. Of these, 22 P/LP variants in 15 genes associated with 13 diseases of 85 individuals (5.5%; 85/1559) were also reported as P/LP in the ClinVar archive. The majority harbored variants in genes related to cardiovascular diseases (4.7%; 74/1559), followed by cancer phenotypes (0.5%; 8/1559). None of the individuals in our cohort harbored biallelic variants in genes responsible for diseases inherited in an autosomal recessive manner. The results would serve as a basis for precision medicine practice at individual and population levels.