Prediction of Protein Biomarkers for Mycobacterium fortuitum using Machine Learning Technique

M. fortuitum is a recognized rapidly growing bacterium known for causing opportunistic infections in humans. Understanding the pathophysiology of M. fortuitum requires analysis of genes and proteins involved in the growth and multiplication of the bacilli. Global proteome analysis of the bacteria was done in triplicate (P1, P2, and P3). Utilizing the proteome dataset of M. fortuitum, neural network (NN) analysis, a machine learning technique, was utilized to find the differential expression profile of proteins for the first time in order to develop a model that aids in the prediction of biomarkers. Moreover, employing NN analysis yielded a prediction model that was more precise and effective than those derived from conventional techniques, underscoring the potential of this approach to promote the identification of biomarkers for infectious illnesses. The mean square error and coefficient of determination were calculated for each of the three data sets. The model which had various levels of protein expression was evaluated. As a consequence of the investigation, the model with greater protein expression levels results lower mean square error values and higher coefficient of determination values. P1 was tested to see if it met the requirements for training data for more combinations using the P2 and P3 sets. This implies that adding additional information on protein expression enhances the prediction model’s ability to accurately identify potential biomarkers in bacterial diseases like M. fortuitum.