Abdullah Hanif

PhageMatch is designed to make phage selection more reliable by predicting phage bacteria interactions with machine learning. It supports decision making when labeled interaction data is limited by learning broader biological patterns and improving generalization to new phage host pairs. PhageMatch integrates multi omics and multi modal signals and models relationships using graph based methods, including graph neural networks and hypergraph modeling. It also uses self supervised learning to extract value from unlabeled data and strengthens robustness with pre trained biological models and ensemble approaches for more consistent predictions. A step toward faster, more precise phage therapy workflows across healthcare and agriculture. #PhageMatch #MachineLearning #GNN #Omics #ComputationalBiology #HealthTech

Antibiotic resistance is accelerating, and more infections are becoming harder to treat with standard antibiotics. This is pushing renewed interest in bacteriophages as a targeted option against specific bacterial pathogens. The challenge is that phage therapy depends on the correct match between a phage and a bacterial strain, and that match is often difficult to identify quickly and reliably. Interaction data is limited, biology is highly diverse, and results that work in one context do not always transfer to another. This gap slows down real world adoption of phage therapy in time sensitive settings. #AntibioticResistance #PhageTherapy #PublicHealth #Microbiology #Bioinformatics