On February 12, a biotechnology company dedicated to developing novel therapies using a unique generative AI platformIambic TherapeuticsWith partners inNature Machine IntelligencePublished an article titled "State-specific protein–ligand complex structure prediction with a multiscale deep generative model"The Latest Paper"

The paper shows that Iambic Therapeutics' first-in-class generative AI model, NeuralPLexer, outperforms other state-of-the-art systems in predicting the structures of protein-ligand complexes and modeling the conformational changes induced by adding drug molecules to these structures.

Binding complexes formed by proteins and small molecule ligands are vital to life. Despite recent advances in protein structure prediction, existing algorithms are currently unable to systematically predict the structures of bound ligands and their regulatory effects on protein folding. To address this challenge, Iambic Therapeutics, Inc. has partnered with collaborators to introduceNeuralPLexer, capable of directly predicting protein-ligand complex structures using protein sequence and ligand molecular graph inputs.

In predicting the accuracy of protein structures for representative structure pairs involving large conformational changes and recently determined ligand-binding proteins, NeuralPLexer consistentlyOutperforms AlphaFold2。

Iambic Therapeutics, Inc. was founded in 2019 and is headquartered in California, committed to disrupting the therapeutic field with its unique AI-driven drug discovery platform. In July 2021, Iambic Therapeutics announced the completion of a $53 million Series A financing round, officially stepping out of stealth mode; in October 2023, the company announced the completion of a $100 million Series B financing round, with investors including the highly regarded NVIDIA. Iambic Therapeutics also made it to the list with its Series B financing.Top 10 Global AI Pharmaceutical Financing in 2023"List."

Iambic Therapeutics is using NeuralPLexer to build the company's pipeline, among which the projects with relatively faster progressIAM1363IAM1363 is a selective, brain-penetrant small molecule inhibitor of wild-type and oncogenic mutant HER2 proteins. It has been specifically designed to avoid the toxicity caused by off-target EGFR inhibition. In preclinical studies, IAM1363 demonstrated over 1000-fold selectivity for HER2 over EGFR and showed promising efficacy and tolerability across various HER2 tumor models. Recently, the FDA has accepted the IND application for IAM1363, with related clinical trials planned to begin in early 2024.

On the same day the new paper was published, Iambic also released a white paper introducing the next generationNeuralPLexer2Technology. NeuralPLexer2 was trained in October 2023, achieving significant progress in prediction accuracy, and the model has been expanded to include most categories of biomolecular structures, including protein-protein complexes, cofactors, post-translational modifications (PTMs), and protein-nucleic acid complexes, covering nearly all structures in the Protein Data Bank (PDB).

Notably, NeuralPLexer2 boasts fast prediction speed.Approximately 50 times faster than AlphaFold2Provides opportunities for large-scale computer screening and research on conformational responses triggered by ligand binding.

Iambic Therapeutics claims that NeuralPLexer2 is expected to revolutionize structure-based drug discovery and may lead a new wave of AI-driven drug discovery.

References:

[1]https://www.iambic.ai/post/iambic-therapeutics-announces-new-research-published-in-nature-machine-intelligence-demonstrating-the-capabilities-of-its-generative-ai-neuralplexer-technology-to-predict-protein-ligand-complex-structures

[2]https://www.nature.com/articles/s42256-024-00792-z

[3]https://www.iambic.ai/post/transforming-computational-drug-discovery-with-neuralplexer2

[4]https://www.iambic.ai/post/her2-ind