Computer Program Enumerates 2 Billion Drug-Like Molecules

Posted on

Harran Houk small 0

Harran and Houk groups create computer software that predicts structure and three-dimensional conformations of products made from synthetic reaction sequences.  

The Composite Peptide Macrocycle Strategy, used to make the synthetic reaction sequences, was perfected in Professor Patrick Harran’s lab. Professor Ken Houk’s group worked with Harran’s group to create the software. 

Houk Harran

The UCLA Team – Professors Patrick Harran and Ken Houk, chemistry graduate student Ishika Saha, postdoctoral fellow Dennis Svatunek, and computer science graduate student Eric Dang.

The research was described in the latest issue of PNAS. Their software CPMG (Composite Peptide Macrocycle Generator), combinatorically generates linear peptides from novel amino acid building blocks, and processes the sequences according to Harran’s multi-step reaction sequences to enumerate products that can be made efficiently in the Harran labs. Three examples with quite different shapes, but which can be made by the Harran methodology, are shown below.


CPMG can enumerate trillions of possible compounds, but which ones of these should be synthesized first? Developing ways to determine that is a future goal of this research project.

Peptide derived macrocycles can bind to challenging protein targets and thus serve as lead structures for drug discovery. The Harran lab synthesizes organic template molecules that are amalgamated with peptides to form composite macrocycles that retain molecular recognition elements in the product. These polycyclic structures are anticipated to have important pharmacological properties. Harran and graduate student, Ishika Saha, teamed up with Eric Dang, a Computer Science graduate student, and Houk and his postdoc, Dennis Svatunek, to create the computational method, CPMG. The UCLA group has now partnered with the Stefano Forli lab at Scripps to predict which of these products are most likely to be ligands for proteins involved in diseases. Through the combination of CPMG and high-throughput docking, only highly promising lead molecules will be synthesized and tested.