Custom search configurations produce better automated synthesis routes

Despite substantial developments in recent decades, synthetic chemistry remains an extremely demanding field. Although creating new molecules with distinct characteristics plays a crucial role in therapeutic discovery, it remains a time-consuming endeavor that can often become a bottleneck in the overall workflow.

Highly experienced synthetic chemists devote significant time and consideration to route scouting and optimizing reaction conditions to access new structures. This dependence on human expertise creates opportunities for errors and biases and is limited by scalability and efficiency. As the demand for novel molecules with specific characteristics continues to grow, the need to streamline the synthesis workflow has become urgent.

SYNTHIA^™ Retrosynthesis Software application displaying the graph view from the search results of a target molecule (left) and Synple^™ automated synthesizer shown with solvent reservoirs (right). Image Credit: Merck KGaA

Enabling technologies

Fortunately, developments in technology and automation are now transforming synthetic chemistry. Today, the rise of robotic systems and enabling tools, machine learning algorithms, and virtual screening techniques are helping to speed up the discovery and synthesis of novel molecules.

These technologies have the potential to decrease the time and effort needed to generate a solution for a given synthesis while also minimizing the risk of errors. By combining the robustness of automation to standardize chemical synthesis with a data-driven AI strategy to quickly analyze large data sets and predict novel results, scientists can investigate a broader chemical space in considerably less time.

SYNTHIA^™ retrosynthesis software, a robust platform for organic chemists to develop synthetic routes, is an excellent example of a technology that aids synthesis design. This software utilizes expert-coded rules to predict feasible routes from commercially available starting materials and can speed up the route design process by providing researchers with a tailored set of ranked potential routes to any given target.

From the perspective of synthesis automation, Synple’s automated cartridge-based synthesis technology is a user-friendly, enabling tool capable of performing many of the chemical reactions commonly employed in discovery chemistry. The device and cartridges allow scientists to initiate and complete the reaction, work up, and product isolation process with the press of a button, resulting in substantial savings in both time and cost.

This case study presents how these two novel technologies can accelerate the design and synthesis processes, while also examining how interfacing them can increase the productivity of this combined process. This combination is achieved by implementing a specific set of search parameters in SYNTHIA^™, enabling it to identify routes that can be effected automatically through the Synple platform.

SYNTHIA^™ search configurations

SYNTHIA^™ offers multiple configurations for performing searches aimed at identifying routes suited to specific applications (Figure 1). In addition to the preset search options, users can develop custom search configurations tailored to their specific project requirements.

This study compares two of the preset search configurations, General and Discovery, with a third custom configuration, Synple, that promotes reactions that can be executed using the Synple automated synthesis system.

Figure 1. SYNTHIA^™ Search Configurations. General balances number of steps with cost of starting materials. Discovery prioritizes shorter routes over inexpensive starting materials to meet the needs of discovery chemists who need to quickly access small amounts of their target. Synple is a custom configuration based on the Discovery setting that prioritizes reactions that can be executed using the Synple automated synthesis system. Image Credit: Merck KGaA

Configuration comparisons

A test set of 10 representative molecules, chosen from known therapeutic candidates and a drug-like library, was utilized to evaluate the output of the different search configurations (Figure 2). All ten molecules were uploaded to SYNTHIA^™ as a single file using the Batch Retrosynthesis Module. The batch was run with each of the different search configurations, which were chosen from a drop-down list during the search setup.

The retrosynthetic evaluations for each target required between 9 and 16 minutes to complete. For each configuration, the highest-ranked pathway was given a score for the total number of synthetic steps (including any necessary protection or deprotection) and the number of steps compatible with execution using Synple (Table 1).

As anticipated, the General configuration typically identified retrosynthetic pathways containing the highest number of steps using lower-cost starting materials. In contrast, the Discovery and Synple configurations prioritized pathways with fewer steps. A more detailed evaluation of the top pathways from the latter two configurations highlighted two notable differences:

1. The Discovery configuration frequently utilized more challenging or exotic reactions, whereas the Synple configuration returned pathways with steps mostly using routine transformations that could also be performed using the Synple synthesis system.
2. Since the Synple configuration was engineered to promote protection/deprotections available as Synple cartridges (for example, Boc), these routes were more likely to use commercially available protected starting materials and included exact deprotection steps. Routes from the Discovery configuration were more likely to include general protection/deprotection steps from unprotected starting materials, providing users with the flexibility to choose from several compatible protecting group options.

Figure 2. Compounds used in batch. C1. AAK1 inhibitor,¹ C2. Antithrombotic agent², C3. Neurokinin antagonist³, C4. elF4A3 inhibitor⁴, C5-7. Library Compounds⁵, C8. CD38 inhibitor⁶, C9. GR antagonist⁷, C10. Library Compound. Image Credit: Merck KGaA

These patterns are illustrated by the analysis of the C3 results, a Neurokinin antagonist candidate.³ The General configuration proposes a six-step route beginning with more cost-effective starting materials and including just a single step that can be automated via the Synple system (Scheme 1).

The top routes for the Discovery and Synple configurations contain five total steps. However, only one is potentially automatable using the Discovery route, while all five steps can be automated in the Synple route. In addition, the Synple route identifies a Boc-protected starting material and includes the Boc-deprotection as an explicit step, while the Discovery route starts from an unprotected starting material and allows users to select the most appropriate protecting group from a range of compatible options.

The more common synthetic transformations proposed in the Synple route may also be more attractive, as they can reduce the likelihood of failure and minimize the need for optimization.

It should be noted that while the retrosynthetic evaluation may be carried out through the Synple configuration, users are still free to choose from 50 possible routes containing both automatable and manual steps.

As an example, the shortest route to reach C8 involved two Synple steps and one manual step, demonstrating an instance where the most optimal route combines the strengths of both the chemist and the machine. By offering the flexibility to create target molecules through a combination of both automated and manual chemistry, SYNTHIA^™ provides chemists with the freedom to automate the simple steps while reserving their skills and energy for critical non-automatable synthetic steps.

Figure 3. Synple customization using Seek Function to prioritize keywords and a list of substructures related to Synple-enabled reaction. Image Credit: Merck KGaA

Table 1. Total number of synthetic steps/Number of steps compatible with Synple for each batch molecule across different search configurations. Source: Merck KGaA

Synthetic execution

To assess the synthetic routes recommended by SYNTHIA^™ and showcase the advantages of interfacing SYNTHIA^™ and Synple synthesis automation, a test molecule was chosen for synthesis using the routes recommended by the Discovery configuration (manual synthesis) and the Synple configuration (automated synthesis).

C6 synthesis using the Discovery route required approximately six hours of “hands-on” working time (excluding the reaction time during stirring). The initial alkylation step required additional scouting for different conditions due to the reactivity of the alkyl halide.

Using the route suggested by the Synple configuration, just one hour of working time was needed to configure the two steps on the Synple machine, followed by solvent evaporation after the automated reaction.

Both routes produced comparable overall yields (49% vs 46%). However, the product from the Synple route was sufficiently pure at 90%, eliminating the need for any further purification.

This article demonstrated the fast and efficient synthesis of the proposed SYNTHIA^™ computed routes using both traditional and automated methods. With the complete workflow, C6 was accessed in just a few hours of working time, with no drawbacks compared to the traditional method.

Summary

Applying a customized set of search parameters in SYNTHIA^™ enables efficient planning of routes to target molecules that can be rapidly executed using Synple automated synthesis technology. This reduces manual effort for route planning and searching for adequate starting materials, as well as bench execution, ultimately speeding up the pharmaceutical discovery process.

Since the Synple automatable chemistry search configuration was adopted in SYNTHIA^™ with relative ease, this application can be extended to prioritize any subset of reactions that individual users would prefer to employ for their projects. As a result, SYNTHIA^™ has the potential to serve as a robust tool for uniquely tailored, focused retrosynthetic evaluation.

References and further reading: 

1. Luo, G. et al. (2015) WO2015153720A1.
2. Priepke, H. et al. (2007) WO2007003536A1.
3. Shue, H.-J. et al. (1996) WO1996034864A1.
4. Kulkarni, S. S. et al. (2021) WO2021087087A1.
5. Ito, M., et al. (2017). Discovery of Novel 1,4-Diacylpiperazines as Selective and Cell-Active eIF4A3 Inhibitors. Journal of Medicinal Chemistry, 60(8), pp.3335–3351. DOI: 10.1021/acs.jmedchem.6b01904. https://pubs.acs.org/jmcmar/article-abstract/60/8/3335/764504/Discovery-of-Novel-1-4-Diacylpiperazines-as?redirectedFrom=fulltext
6. McMillan, A. E., et al. (2022). A vending machine for drug-like molecules – automated synthesis of virtual screening hits. Chemical Science, 13(48), pp.14292-14299. DOI:10.1039/d2sc05182f. https://pubs.rsc.org/sc/article/13/48/14292/786792/A-vending-machine-for-drug-like-molecules
7. Rimland, J., et al. (2010). The identification of a novel, selective, non-steroidal, functional glucocorticoid receptor antagonist. Bioorganic & Medicinal Chemistry Letters, 20(7), pp.2340-2343. DOI:10.1016/j.bmcl.2010.02.019. https://www.sciencedirect.com/science/article/abs/pii/S0960894X10001551

About Merck KGaA

Merck Millipore’s purpose is to solve the toughest problems in life science by collaborating with the global scientific community – and through that, we aim to accelerate access to better health for people everywhere.

We provide scientists and engineers with best-in-class lab materials, technologies and services. With the 2015 combination of Merck Millipore and Sigma-Aldrich, we now have a broad portfolio of 300,000 products, an expanded global footprint and an industry-leading eCommerce platform - SigmaAldrich.com.

We are dedicated to making research and biotech production simpler, faster and safer.

Our three life science business areas are designed to best address our customers’ needs:

Research Solutions: Serve customers focused on identifying and developing medicines.

Process Solutions: Deliver end-to-end products and expertise to customers who take what is developed in labs and manufacture it.

Applied Solutions: Support customers in ensuring that drugs, food and beverages are safe for consumption.

Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.net, which is to educate and inform site visitors interested in medical research, science, medical devices and treatments.