PREDICTING GENETIC VARIANTS PATHOGENECITY

Computer Engineering

Single Nucleotide Polymorphism (SNP) detection plays a pivotal role in understanding the intricacies of genetic inheritance and its influence on phenotypic traits. This project focuses on developing a predictive model that analyzes and determines the parental origin of specific SNPs in a child’s DNA. By leveraging genetic data from both parents, we aim to gain deeper insights into hereditary patterns and how they contribute to observable characteristics. To accomplish this, we implemented a logistic regression model that integrates not only genetic information but also lifestyle and environmental factors such as smoking, alcohol consumption, radiation exposure, and mutation scores. These factors were included to examine their potential impact on SNP expression and inheritance. The model was trained and tested on a dataset composed of these variables, resulting in an overall accuracy of 58%, an ROC AUC of 0.604, a recall of 63.27%, and a precision of 56.36%. These metrics suggest the model has moderate predictive capabilities but also emphasize the need for more sophisticated feature selection and machine learning algorithms to enhance performance. In conclusion, this project underlines the multifaceted nature of genetic prediction and the influence of environmental variables on genetic traits. While the logistic regression model provides a foundational approach to SNP inheritance analysis, future work will explore advanced models and deeper biological data integration to improve accuracy and practical applicability in genetic research and personalized medicine.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.