DOMINIK SARMA
Belyntic GmbH, Berlin, Germany

Abstract

Artificial intelligence (AI) breakthroughs will significantly impact the development of new medicines. For example, AI can help decipher a protein’s three-dimensional structure providing a rich repository for therapeutic hits. This trend holds considerable potential for peptide drugs because they play a crucial role in protein-protein interaction. AI is also used in the field of cancer and infectious diseases to find T-cell regulatory peptides.
To unfold this potential, however, developers must validate and improve the therapeutic effect of the predicted peptides calling for synthetic high-throughput solutions and complex chemical modifications. Furthermore, manufacturers must reduce the environmental impact of complex peptide synthesis. A holistic process design that implements automation, digitalization, and sustainability can help solve these challenges to sustain peptides as a viable drug class in the future.

The release of Google’s deep-learning tool Alpha Fold on November 30, 2020, stands for a paradigm shift in pharmaceutical R&D. Today, researchers can rapidly predict a protein’s 3D shape from its amino acid sequence and design molecules capable of modulating the protein function using artificial intelligence (AI). Moreover, this knowledge is made available to the public in an open-access platform that includes structural information for the human proteome and 20 other vital organisms (1). Besides the enormous potential to discover novel therapeutic agents, the increasing attention of computer-assisted R&D is also an important measure to counteract declining returns and increasing cost and development times (2). As a result, new drug leads will be more and more the result of computational power rather than extensive research and chemical or biological screening programs in the future.

This trend holds considerable potential for peptide-based drugs, as they play a crucial role in, for example, protein-protein interactions (PPIs). With the advanced knowledge about peptide modifications to overcome their inherent therapeutic drawback, e.g., in-vivo ...