A team of researchers from the Indian Institute of Technology Madras (IIT-Madras) has created a computational tool using machine learning (ML) to enhance the detection of cancerous tumours in the brain and spinal cord. Named ‘GBMDriver’ (GlioBlastoma Multiforme Drivers), this tool is now accessible to the public through an online platform.
Glioblastoma is a highly aggressive tumour that grows rapidly in the brain and spinal cord. Despite research efforts to comprehend this tumour, treatment options remain limited, and the prognosis is typically poor, with a survival rate of less than two years from the initial diagnosis.
Evaluating the functional consequences of protein variants associated with Glioblastoma is crucial for advancing therapeutic options for patients. However, conducting functional validations to identify driver mutations, which are the specific mutations responsible for causing the disease, from the multitude of observed variants would be laborious work.
According to a statement by IIT-Madras, the GBMDriver tool was specifically designed to identify driver mutations, which are responsible for the development of Glioblastoma, and passenger mutations, which are neutral mutations. During the development of the web server, multiple factors were considered, including amino acid properties, di- and tri-peptide motifs, conservation scores, and Position Specific Scoring Matrices (PSSM).
In this study, 9,386 driver mutations and 8,728 passenger mutations in glioblastoma were analysed. Driver mutations in glioblastoma were identified with an accuracy of 81.99%, in a blind set of 1,809 mutants, which is better than existing computational methods. This method is completely dependent on protein sequence, the statement explained.
- Michael Gromiha from the Department of Biotechnology at IIT-Madras provided insights into the key findings of the team’s research. He stated that they had successfully identified crucial amino acid features for the identification of cancer-causing mutations and achieved the highest level of accuracy in differentiating between driver and neutral mutations. The team’s aim is for the GBMDriver tool to aid in prioritising driver mutations in glioblastoma and support the identification of potential therapeutic targets. Ultimately, this tool can contribute to the development of effective drug design strategies for the treatment of glioblastoma.
The research findings hold several key applications, including:
- The methodology and features employed in this research can be adapted and applied to other diseases beyond glioblastoma.
- The method developed in this research can serve as a significant criterion for predicting disease prognosis. By accurately identifying driver mutations, it can aid in understanding the severity and progression of the disease, contributing to improved patient management and personalised treatment approaches.
- The research provides a valuable resource to identify mutation-specific drug targets to design therapeutic strategies.
Last year, IIT-Madras researchers developed PIVOT, an artificial intelligence (AI)-based tool that can predict cancer-causing genes. As OpenGov Asia reported, PIVOT predicts cancer-causing genes using a model that utilises information on mutations, expression of genes, and copy number variation in genes and perturbations in the biological network that results from an altered gene expression. The tool applies machine learning to classify genes as tumour suppressor genes, oncogenes, or neutral genes. PIVOT successfully predicted both the existing oncogenes and tumour-suppressor genes like TP53, and PIK3CA, among others, and new cancer-related genes such as PRKCA, SOX9, and PSMD4.