Mutation types and pathogenicity classification using multi-label multi-class deep networks.

In recent years, considerable progress has been made in genomics and proteomics, resulting in much biological data. To draw inferences from this data, advanced computer analysis techniques are required. Bioinformatics is crucial to interpreting and applying this data in any endeavor. As huge amounts of genomic, proteomic, and other data begin to be collected and integrated, the relevance of this emerging area of study will increase. The detection of variation in Next Generation Sequencing is now a standard and indispensable technique in all areas of the biological sciences. Life-threatening conditions like cancer and other diseases may be caused by DNA mutation sequences (genetic mutations). Thus, it's important to discover these mutations early, classify them, and understand their effects on the DNA sequence. Bioinformatics is mostly concerned with the changes to DNA that will occur in the cells of the next generation. The primary goal of this research is to develop a sophisticated and comprehensive approach for classifying the mutation type and categorizing the influence of known disease-causing genetic variations on an individual's risk of disease using a multi-label multi-class deep learning classification technique. A deep neural network was trained using the TP53 dataset (The TP53 gene is a tumor suppressor gene on chromosome 17) to discover the mutation type and its impact in a unified model, as described. To evaluate the performance of the suggested method, the accuracy, recall, precision, and F1 score evaluation scales were utilized. and the results demonstrate that it can accurately determine the mutation type and its effects with an accuracy of 97.58%, Precision 98%, Recall 95%, and F1 score 96%. [ABSTRACT FROM AUTHOR]