This repository contains the code to reproduce results in:

A Machine Learning approach for ACMG/AMP-based Genomic Variant Classification and Prioritization by G. Nicora, S. Zucca, I. Limongelli, P. Magni and R. Bellazzi.

Classification and prioritization of genomic variants associated with inherited disorders.

inherited variants are nowadays interpreted according to the ACMG/AMP guidelines [1], which define 5 different tiers of classification: Pathogenic, Likely pathogenic, Benign, Likely benign, VUS (variant of uncertain significance). Therefore, when evidence is not enough for classification, a variant can be still uncertain (VUS). To interpret VUS variants, data-driven approaches, such as Machine Learning algorithms, can be used.

Variants are annotated and interpreted according to the ACMG/AMP guidelines [1] by the eVai software [2]. ACMG/AMP levels of evidence are used as features for Machine learning approaches (in particular, Logistic Regression).

We applied our framework on different datasets. Results are shown in the Notebook.

[1] Richards, S., Aziz, N., Bale, S. et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 17, 405–423 (2015). https://doi.org/10.1038/gim.2015.30

[2] https://www.engenome.com/wp-content/uploads/2020/10/eVai_enGenome_WhitePaper_v0.7.pdf