Enzymes are nature's own biocatalysts and are involved in almost all the processes in the cycle of life in all biological systems. However, despite millions of proteins being discovered, less than 2% have well-known functions. This means that the majority of the proteins that are sequenced at an exponential scale is yet to be functionally characterized.
Centre for Young Synbio Scientists doctoral student Robert Armah-Sekum is looking into ways to leverage AI and computational modeling to help predict the functions of yet-to-be-characterized proteins.
Proteins fold up into specific shapes according to the sequence of amino acids in the polymer, and the protein function is directly related to the resulting 3D structure. Understanding this relationship would help predict the function of the sequenced proteins without the need for time-consuming trial and error experimental testing.
The research aims to specifically look into enzymes and ways of predicting their catalytic activity and substrate interaction from different protein attributes. Robert Armah-Sekum and his team are using a machine learning model to learn from previous experimental data from protein attributes such as the structure of proteins, the sequence nature of proteins, and how well some of the proteins align to each other.
While previous methods focus mostly on the input space of protein features, their approach takes into consideration challenges on the output side to boost the prediction accuracy and speed of the task. The machine learning model that they are using for this prediction task is designed to run in a relatively faster time compared to previous methods and doesn't take up much space on the computer.
Combining computing technologies with synthetic biology promises to accelerate the development of new therapies and also helps industries adapt to more sustainable production and manufacturing. Robert Armah-Sekum says that he picked up the interest in this area during his undergraduate studies in biomedical engineering. He notes that there is a clear synergy between computing technologies and synthetic biology that can create revolutionary things.
The research conducted by Robert Armah-Sekum contributes to the following UN sustainability goals:
Goal 3: Good Health and Well-being - The research aims to accelerate the development of new therapies through the prediction of enzyme functions. This has the potential to lead to the discovery of novel treatments and medical solutions, advancing global health and well-being.
Goal 9: Industry, Innovation, and Infrastructure - By leveraging AI, computational modeling, and machine learning to predict enzyme functions, this research aligns with the goal of fostering innovation and enhancing scientific infrastructure, which can lead to more efficient and sustainable industrial processes.
Goal 12: Responsible Consumption and Production - Predicting enzyme functions using computational methods can reduce the need for time-consuming experimental testing, thereby contributing to more responsible consumption of resources in the research and development process.
Goal 13: Climate Action - The use of computational technologies and synthetic biology to accelerate research and development can lead to more efficient and sustainable production and manufacturing processes, potentially reducing the environmental impact of various industries.
Goal 17: Partnerships for the Goals - The collaborative approach of combining computing technologies with synthetic biology demonstrates a partnership between different fields of expertise to achieve sustainable and innovative outcomes.