Three out of every four new or emerging infectious diseases in humans can be traced back to animals according to the Centers for Disease Control and Prevention (CDC).
In a new study published in Cell Reports, scientists have applied artificial intelligence (AI) machine learning to help predict which viruses might spread from animals to humans—an advancement that may one day help prevent viral infections and pandemics.
The team applied AI deep learning and created a graph convolutional neural network (GCNN) to learn a representation for glycans and called their model SweetNet.
Graph convolutional neural networks are used for a variety of purposes such as computer vision, bioinformatics, traffic prediction, fraud detection, and social analysis.
“SweetNet is a graph convolutional neural network that uses graph representation learning to facilitate a computational understanding of glycobiology,” wrote the scientists.
The researchers found that the glycan representations learned by SweetNet outperformed alternative models.
Specifically, the SweetNet-based models of glycans require less training time and have greater data efficiency.
By applying AI deep learning to glycobiology, scientists have taken an important step towards using glycans as potential antivirals to prevent viral infections and pandemics in the future.
- Why getting digital transformation fundamentals right keeps customers happy
- How Conversational AI Works and What It Does
- Creatio: How No-Code Automation Drives Organizational Change
- How The Digital Age Is Reinventing (Almost) Everything
- DALL·E: Creating Images from Text
- OpenAI’s system for translating plain English into code looks impressive — and slightly scary
- Integration and automation are mission critical for connection