AI and chemical simulation-based drug discovery support service
drug2drugs® is a service to provide you multiple active compounds with different scaffolds from only the structural information of one active compound without knowing the target protein, so that we can increase the overall success rate of clinical drug development.
An active compound is a compound that has the ability to bind to its target protein. Having multiple active compounds with diverse scaffolds is important for improving the success rate in drug development to avoid various problems, such as low solubility, high toxicity or low permeability.
Fujifilm has developed AI-AAM® (AI-Amino-Acid Mapping), a new scaffold transformation method based on AI and chemical simulation, with the aim of increasing the number of active compounds with various scaffolds from one active compound. Scaffold transformation using AI-AAM® requires only the structural information of an active compound, not the information of the target protein.
We start drug2drugs®, a service that increases the number of active compounds from active small molecules or peptides by scaffold transformation of AI-AAM®.
How to increase active compounds with AI-AAM®
AI-AAM® that consisting of AAM descriptors and AI structure generator
1. AAM (Amino-Acid Mapping) descriptors[1]
AAM descriptors are the set of the 3D existence probability map of 20 amino acids around each compound computed by chemical simulations. Two compounds with the same AAM descriptors but different scaffolds interact with the same amino acids and bind to proteins with the same binding pattern.
2. De novo design by our AI[2]
Our developed AI automatically designs chemical structures with completely different scaffolds but equivalent AAM descriptor. Since the initial structure is not required and the design is performed at the atomic level, about 1060 compounds can be designed. The compounds designed by our AI exhibit high thermal stability.
Experimental results of scaffold hopping of active compounds of a small molecule or a peptide using AI-AAM®
1. Scaffold hopping between small molecules[3]
From the structural information of BIIB-057, XC608 was selected as the compound with the most similar AAM from approximately 12 million compounds in our chemical library within a month.
2. Scaffold hopping from a peptide to a small molecule[4]
From the binding conformation and site of the Peptide-71, the partial structure of Peptide-71 in contact with PD-L1 when binding to the protein, the small molecule compound FF-ATC-001 was also designed within a month.
| [1]JP6826672B2, [2]JP7191969B2, JP7190498B2, JP7116186B2, [3] CBI Annual Meeting 2022, P08-14; bioRxiv (2023), doi:https://doi.org/10.1101/2023.07.03.547598, [4] CBI Annual Meeting 2022, P08-07, Excellent Poster Awards, [5]WO 2009/136995, [6]WO 2011/035077, [7]US20140294898 A1 |
Correlations between binding affinity and AAM similarity
The AAM similarities of 22 compounds computed from the reference SYK inhibitor BIIB-057 exhibited the high correlation with the target affinity.
References
A novel in silico scaffold-hopping method for drug repositioning in rare and intractable diseases
Tanabe, M et al. : bioRxiv (2023)
A service that increases the number of active compounds using AI-AAM®
Here is the basic procedure of our drug2drugs® service. You should provide us the structural formula of the compound of the interest in the case of scaffold hopping between small molecules, and the binding conformation and site of your peptide for the scaffold hopping from a peptide to small molecules. Using the information, we first search from our library of the real chemicals and propose promising compounds. You then purchase the compound, evaluate its activity, and inform us of the result without letting us know the target information. Next, we use the results to design new compounds and propose more promising structures.
For research use or further manufacturing use only. Not for use in diagnostic procedures.
Product content may differ from the actual image due to minor specification changes etc.
If the revision of product standards and packaging standards has been made, there is a case where the actual product specifications and images are different.