South High Community School
Infectious Disease and Microbiome Program
The fungal pathogen Cryptococcus neoformans threatens many immunocompromised individuals worldwide. Afua’s project centered on discovering new drugs that could inhibit the growth of this fungus. Afua used previously-collected experimental data showing the effectiveness of thousands of drug candidates at inhibiting fungal growth to train a machine learning model, based on a neural network framework. She then tested the model’s accuracy, and found that it was 99% likely to correctly classify compounds from the training set as “effective” or “ineffective.” Afua then analyzed the chemical structures of some of the most effective compounds, and found structural motifs that were common to all of the active compounds, resulting in a subset of compound structures that will be used for future compound development. “BSSP was a really appealing program for me because it promised to provide the opportunity to do research,” Afua said. “The scientific talks and social events in the program would allow me to both explore different areas of science that I am interested in, as well as meet other scientists who were in my shoes not too long ago!” Afua's favorite part of being a Broadie was the community and the endless opportunities. "At Broad, I feel welcome and I feel like I can freely receive and give support. I feel like I never have to do anything alone; everyone is always eager to help out no matter how big or small. I am very proud to be part of this amazing community."