Automation of route identification and optimisation based on data-mining and chemical intuition

Abstract

Data-mining of Reaxys and network analysis of the combined literature and in-house reactions set were used to generate multiple possible reaction routes to convert a bio-waste feedstock, limonene, into a pharmaceutical API, paracetamol. The network analysis of data provides a rich knowledge-base for generation of the initial reaction screening and development programme. Based on the literature and the in-house data, an overall flowsheet for the conversion of limonene to paracetamol was proposed. Each individual reaction-separation step in the sequence was simulated as a combination of the continuous flow and batch steps. The linear model generation methodology allowed us to identify the reaction steps requiring further chemical optimisation. The generated model can be used for global optimisation and generation of environmental and other performance indicators, such as cost indicators. However, the identified further challenge is to automate model generation to evolve optimal multi-step chemical routes and optimal process configurations.

Statistics of the network of organic chemistry

“…9 In our own work, we have shown the use of large chemical datasets to develop reaction sequences by running a targeted network search, taking molecular structural information into account; the reaction sequences are then evaluated in terms of a range of performance metrics. 12,13 In addition to synthesis planning, an alternative potential use of chemical data networks is the discovery of new reac-tions. This is an inverse problem: instead of asking a chemical question from the network, we intend to ask a mathematical question, with a hypothesis that the structure of the chemical network contains implicit chemical information, which we may reveal in the form of yet unknown transformations.…”

The Molecular Industrial Revolution: Automated Synthesis of Small Molecules

“…[56–58,60–62] As such algorithms and broad access to the required computing power continue to advance, and new opportunities opened up by recent advances in artificial intelligence/machine-learning are leveraged, it is likely that a substantial portion of such customized synthetic planning can be achieved automatically. [60,62,63] …”

“…[56][57][58][60][61][62] As such algorithms and broad access to the required computing power continue to advance and as new opportunities opened up by recent advances in artificial intelligence/machine-learning are lever-aged, it is likely that asubstantial portion of such customized synthetic planning can be achieved automatically. [60,62,63] Ap otentially even more challenging problem associated with this approach is that it requires creating am achine that can perform as many different types of reactions as possible,is compatible with many different types of starting materials and reagents,a nd can execute many different types of purifications.E ach of these goals alone is challenging;i n combination, they represent af ormidable task.…”

Mapping supply dynamics in renewable feedstock enabled industries: A systems theory perspective on ‘green’ pharmaceuticals

“…7 Multi-stage multi-layer mapping methodology for industrial network systems enabled by renewable chemical feedstocks respectively. Thereafter, the future of bio-based supply networks relies on either the technical capability to integrate novel synthesis routes to existing infrastructure and processing facilities (Lapkin et al 2017), or the economic feasibility of investments in new biorefineries (Gunukula et al 2018). Especially, commercialisation of new biopharmaceutical molecules requires consideration of key technoeconomic parameters (Mupondwa et al 2015), including: medications' biological and pharmacological potency, capital investment and operating costs, market uncertainty analysis, production scalability and profitability.…”