Deep Learning for Prediction and Optimization of Fast-Flow Peptide Synthesis

• DOI: 10.1021/acscentsci.0c00979
• Publication Date: 2020-11-12T17:53:35Z
• Authors: Somesh Mohapatra; Nina Hartrampf; Mackenzie Poskus; Andrei Loas; Rafael Gómez-Bomb...; Bradley L. Pentelute
• Keywords: Peptide bond; Amide; Peptide Synthesis; Minification; Sequence (biology)

The chemical synthesis of polypeptides involves stepwise formation amide bonds on an immobilized solid support. high yields required for efficient incorporation each individual amino acid in the growing chain are often impacted by sequence-dependent events such as aggregation. Here, we apply deep learning over ultraviolet–visible (UV–vis) analytical data collected from 35 427 fluorenylmethyloxycarbonyl (Fmoc) deprotection reactions performed with automated fast-flow peptide synthesizer. integral, height, and width these time-resolved UV–vis traces indirectly allow analysis iterative coupling cycles resin. computational model maps structural representations acids sequences to experimental parameters predicts outcome less than 6% error. Our deep-learning approach enables experimentally aware design prediction Fmoc efficiency minimization aggregation events, building foundation real-time optimization flow.