Investigating Potential Natural Antibiotics Plants Based on Unani Formula Using Supervised Network Analysis and Machine Learning Approach

This study employs a multi-faceted approach to identify potential natural antibiotics derived from plants used in the Unani formula. The first approach involves a supervised network analysis that utilizes distance measurement to analyze the interconnectivity of the network data. This approach identifies 26 candidate plants with the potential for natural antibiotics. The second approach involves machine learning techniques, specifically deep learning methods, to extract relevant features from the Unani data. This method produces a list of 29 potential plants that exhibit characteristics of natural antibiotics. Notably, seven plants overlap between the approaches - Piper longum, Trachyspermum ammi carum copticum, Santalum album, Cyperus rotundus, Vitis vinifera, Matricaria chamomilla and Zingiber officinale - shown in the literature to exhibit antibacterial properties direct or indirect. The study employs two analytical methods to comprehensively identify potential plants with natural antibiotic properties from the Unani formula. These findings could have significant implications for developing novel antibiotics to combat antibiotic-resistant pathogens.