Gregor Mendel's pea plant experiments laid the groundwork for understanding genetics and inheritance. DNA, composed of four bases (A, T, C, G), forms the genetic code, with variants (previously known as mutations) that may be benign, pathogenic, or uncertain (VUS). Humans inherit genes in pairs, with alleles often being dominant or recessive. Medical genetics involves counseling and diagnosing inherited disorders, using tools like DNA sequencing. Bayes’ theorem is key in genetic counseling, updating the probability of genetic events with new evidence and aiding in risk assessment. This method, combined with the American College of Medical Genetics and Genomics (ACMG) guidelines, helps classify genetic variants accurately, assisting in diagnoses. Bioinformatics tools are used to analyze genomic variants across millions of sites, with ACMG criteria guiding their prioritization and reclassification. The Bayesian framework ensures accurate interpretation, crucial for clinical decision-making. A significant challenge in clinical genetics is dealing with VUS, especially in underrepresented populations like Hispanics. This underrepresentation of Hispanic populations in genomic research affects variant classification. The MexVar platform, integrating data from the Mexican Biobank, addresses this by providing a national reference for genomic variation. Through this article we emphasize the mathematical foundation in genetics from the application of Bayes’ theorem to statistical models such as Lasso or Joint-Lassosum. It highlights the challenge of adapting these methods to enhance the predictive accuracy of polygenic risk scores (PRS) for the Mexican population. This is crucial for improving the relevance and precision of genomic insights, enabling better personalized medicine.