Novel biological Gadolinium contrast agents for MRI developed using artificial intelligence

Motivation: MRI contrast agents based on Gadolinium binding peptides with high selectivity for specific targets can increase imaging precision and minimize potential toxicity. Goal(s): Develop novel Gadolinium binding motifs that can be used for molecular imaging. Approach: Peptides were incubated with Gadolinium and MRI scanned to quantify their MRI contrast. The results were used as a dataset for POET (an AI algorithm), which iteratively optimized the peptide sequences for Gadolinium-binding. Results: After two cycles of POET, we found 25 peptides with higher contrast than the initial dataset. The best peptide had 24.3% improvement. The peptides are entirely different than peptides in the initial dataset. Impact: The Gadolinium binding motifs found in this research will be used to develop biological MRI contrast agents that will allow safer and more effective imaging procedures.