(673b) First Principles and Cheminformatics Based in-Silico Tools for Physical Property Prediction Towards Informed Process Development

First-principles and cheminformatics based in-silico tools have recently shown great promise to accelerate drug development pipelines^1-5. Using computational approaches for physical property predictions renders great utility both in early stages of molecular discovery and later stages of product formulation and process development. Solubility is one of the critical properties for design of various unit processes and operations including chemical reactions, extractions and more importantly crystallization processes for synthesis and purification. To characterize the performance of the various unit operations, it is often necessary to measure solubility of a compound in multiple solvents and mixture of solvents as a function of temperature. Thus, many solubility evaluations are required to enable multiple synthetic and physical transformations required to make active pharmaceutical ingredient (API). In this work we present a computational solubility prediction tool, SOLUP, that brings both cheminformatics and first principles-based approaches together in a single platform, where cheminformatics is utilized to improve first principles-based predictions. We will present case studies depicting use of this tool to accelerate the process design and development. In addition to solubility, we will also be talking about case studies where other first-principles based property calculations, such as solvent/API activity, pka, partition coefficient etc., were used to make process decisions . These examples will elucidate the utility of these computational tools to reduce early-stage uncertainty and providing directional knowledge to be used for optimization.

References:

1. Journal of the American Chemical Society 2021 143 (42), 17479-17491
2. Nature Scientific Reports 10, 11492 (2020)
3. AAPS PharmSciTech 23, 18 (2022)
4. Journal of Cheminformatics 13, 98 (2021)
5. Nature Communications 11, 5753 (2020)

Disclosure: Sharad Maheshwari, Eric J. Murphy, James Marek, Ryan G. Ellis, Akshay Korde, Manish S. Kelkar, Daniel A. Pohlman, Richard S. Hong, Nathan S. Abraham, Rajni M. Bhardwaj, Jeremy Henle and Nandkishor Nere are employees of AbbVie. Kushal Sinha and Moussa Boukerche are former AbbVie employees. All authors may own AbbVie stock. AbbVie sponsored and funded the study; contributed to the design; participated in the collection, analysis, and interpretation of data, and in writing, reviewing, and approval of the final publication.