Identifying general reaction conditions by bandit optimization.

Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries1-6. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models7-10 to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.

A Multi-Objective Active Learning Platform and Web App for Reaction Optimization.

We report the development of an open-source experimental design via Bayesian optimization platform for multi-objective reaction optimization. Using high-throughput experimentation (HTE) and virtual screening data sets containing high-dimensional continuous and discrete variables, we optimized the performance of the platform by fine-tuning the algorithm components such as reaction encodings, surrogate model parameters, and initialization techniques. Having established the framework, we applied the optimizer to real-world test scenarios for the simultaneous optimization of the reaction yield and enantioselectivity in a Ni/photoredox-catalyzed enantioselective cross-electrophile coupling of styrene oxide with two different aryl iodide substrates. Starting with no previous experimental data, the Bayesian optimizer identified reaction conditions that surpassed the previously human-driven optimization campaigns within 15 and 24 experiments, for each substrate, among 1728 possible configurations available in each optimization. To make the platform more accessible to nonexperts, we developed a graphical user interface (GUI) that can be accessed online through a web-based application and incorporated features such as condition modification on the fly and data visualization. This web application does not require software installation, removing any programming barrier to use the platform, which enables chemists to integrate Bayesian optimization routines into their everyday laboratory practices.

Implementation of a mathematical model for the photochemical kinetics of a solid form active pharmaceutical ingredient.

We present here the development of a photochemical model used to quantify the risk to photodegradation of a solid drug substance. A key feature of the proposed model development is streamlined estimation of the dependence of the absorption spectra and the quantum yield to the wavelength. A mathematical description of the relationship between the quantum yield and the wavelength enables estimation of photodegradation kinetics under any light anticipated to be encountered in the manufacturing environment. The system studied here consisted of a first order irreversible transformation (A → B(1Φ)) and the formalism strongly suggested the quantum yield was constant over the relevant wavelength range. The predictive power of the model enabled the design of a control strategy to limit the formation of the photo-degradant to very low levels. Also presented are insights obtained from quantum mechanical modeling of the electronic transitions associated with the UV absorption spectra.

A photochemical kinetic model for solid dosage forms.

Photochemical kinetic models to describe the solution phase degradation of pharmaceutical compounds have been extensively reported, but formalisms applicable to the solid phase under polychromatic light have not received as much attention. The objective of this study was to develop a mathematical model to describe the solid state photodegradation of pharmaceutical powder materials under different area/volumetric scales and light exposure conditions. The model considered the previous formalism presented for photodegradation kinetics in solution phase with important elements applied to static powder material being irradiated with a polychromatic light source. The model also included the influence of optical phenomena (i.e. reflectance, scattering factors, etc.) by applying Beer-Lambert law to light attenuation, including effects of powder density. Drug substance and drug product intermediates (blends and tablet cores) were exposed to different light sources and intensities. The model reasonably predicted the photodegradation levels of powder beds of drug substance and drug product intermediates under white and yellow lights with intensities around 5-11kLux. Importantly, the model estimates demonstrated that the reciprocity law for photoreactions was held. Further model evaluation showed that, due to light attenuation, the powder bed is in virtual darkness at cake depths greater than 500µm. At 100µm, the photodegradation of the investigated compound is expected to be close to 100% in 10days under white fluorescent halophosphate light at 9.5kLux. For tablets, defining the volume over exposed surface area ratio is more challenging. Nevertheless, the model can consider a bracket between worst and best cases to provide a reasonable photodegradation estimate. This tool can be significantly leveraged to simulate different light exposure scenarios while assessing photostability risk in order to define appropriate control strategy in manufacturing.

Multivariate Analysis and Statistics in Pharmaceutical Process Research and Development.

The application of statistics in pharmaceutical process research and development has evolved significantly over the past decades, motivated in part by the introduction of the Quality by Design paradigm, a landmark change in regulatory expectations for the level of scientific understanding associated with the manufacturing process. Today, statistical methods are increasingly applied to accelerate the characterization and optimization of new drugs created via numerous unit operations well known to the chemical engineering discipline. We offer here a review of the maturity in the implementation of design of experiment techniques, the increased incorporation of latent variable methods in process and material characterization, and the adoption of Bayesian methodology for process risk assessment.