Author Correction: Controlling an organic synthesis robot with machine learning to search for new reactivity.

Change history: Owing to the misidentification of compound 22 in the original Letter, changes have been made to Fig. 5, Extended Data Fig. 2 and the main text; see accompanying Amendment.

Controlling an organic synthesis robot with machine learning to search for new reactivity.

The discovery of chemical reactions is an inherently unpredictable and time-consuming process1. An attractive alternative is to predict reactivity, although relevant approaches, such as computer-aided reaction design, are still in their infancy2. Reaction prediction based on high-level quantum chemical methods is complex3, even for simple molecules. Although machine learning is powerful for data analysis4,5, its applications in chemistry are still being developed6. Inspired by strategies based on chemists' intuition7, we propose that a reaction system controlled by a machine learning algorithm may be able to explore the space of chemical reactions quickly, especially if trained by an expert8. Here we present an organic synthesis robot that can perform chemical reactions and analysis faster than they can be performed manually, as well as predict the reactivity of possible reagent combinations after conducting a small number of experiments, thus effectively navigating chemical reaction space. By using machine learning for decision making, enabled by binary encoding of the chemical inputs, the reactions can be assessed in real time using nuclear magnetic resonance and infrared spectroscopy. The machine learning system was able to predict the reactivity of about 1,000 reaction combinations with accuracy greater than 80 per cent after considering the outcomes of slightly over 10 per cent of the dataset. This approach was also used to calculate the reactivity of published datasets. Further, by using real-time data from our robot, these predictions were followed up manually by a chemist, leading to the discovery of four reactions.

Publisher Correction: Controlling an organic synthesis robot with machine learning to search for new reactivity.

The chemical structure formatting in Fig. 5 has been corrected online.

Preparation of acetals from aldehydes and alcohols under basic conditions.

A new, simple protocol for the synthesis of acetals under basic conditions from non-enolizable aldehydes and alcohols has been reported. Such reactivity is facilitated by a sodium alkoxide along with a corresponding trifluoroacetate ester, utilizing formation of sodium trifluoroacetate as a driving force for acetal formation. The usefulness of this protocol is demonstrated by its orthogonality with various acid-sensitive protecting groups and by good compatibility with functional groups, delivering synthetically useful acetals complementarily to the synthesis under acidic conditions from aldehydes and alcohols.

Exploration of the Chiral Recognition of Sugar-Based Diindolylmethane Receptors: Anion and Receptor Structures.

In this study, we have conducted a systematic investigation of the chiral recognition of carboxylic anions by D-glucuronic acid/diindolylmethane receptors. We investigate the influence of the anion structure on chiral recognition in the diindolylmethane/glucuronic acid-based receptor 1 a. We found that presence of an additional hydrogen-bond donor at the α position to the carboxylic function is essential for effective chiral differentiation in these systems. Furthermore, we present a synthetic procedure that allows for the synthesis of sugar-decorated receptors that possess a modified substituent at the anomeric position. Four new receptors 1 b-e have been synthesized, and their chiral-discrimination ability toward model carboxylates is studied. The obtained results show that the chiral recognition of these receptors can be fine-tuned by incorporation of a proper substituent into the receptor structure.

Artificial neural networks for guest chirality classification through supramolecular interactions.

A novel strategy for classification of guest chirality based on the combination of artificial neural networks and anion-receptor chemistry is reported. The receptor reported herein forms supramolecular complexes with a variety of biologically important carboxylates, in which the chemical shift changes during addition of anions result in complex guest-stereochemistry-dependent patterns as followed by (1) H NMR spectroscopy. The neural network had learnt these patterns from a training set of 12 anions, and successfully identified the "unknown" chirality of 14 guests present in the test set. Additionally, principal component analysis could discriminate most of the guests studied (26) and allowed for identification of the receptor protons, which are responsible for information transfer of guest chirality.

Bispyrrolylbenzene anion receptor: from supramolecular switch to molecular logic gate.

We have designed anion receptor 4 based on a conformationally labile bispyrrolylbenzene framework, the conformation of which can be changed by appropriate anionic stimuli. In the absence of fluoride anion, the pyrrole moieties rotate freely at room temperature. However, when the concentration of fluoride anion exceeds 2 equivalents, the rotation of the pyrrole units slows down and the conformation of the receptor changes to anti-anti. DFT calculations have shown that this change is due to binding of a third fluoride anion through C-H interaction. Anion receptor 4 can also serve as a molecular logic gate. Anionic inputs such as fluoride and dihydrogenphosphate allow the realization of INHIBIT and NAND logic gate functions with absorption and fluorescence as readouts, respectively.

Electron density-based GPT for optimization and suggestion of host-guest binders.

Here we present a machine learning model trained on electron density for the production of host-guest binders. These are read out as simplified molecular-input line-entry system (SMILES) format with >98% accuracy, enabling a complete characterization of the molecules in two dimensions. Our model generates three-dimensional representations of the electron density and electrostatic potentials of host-guest systems using a variational autoencoder, and then utilizes these representations to optimize the generation of guests via gradient descent. Finally the guests are converted to SMILES using a transformer. The successful practical application of our model to established molecular host systems, cucurbit[n]uril and metal-organic cages, resulted in the discovery of 9 previously validated guests for CB[6] and 7 unreported guests (with association constant Ka ranging from 13.5 M-1 to 5,470 M-1) and the discovery of 4 unreported guests for [Pd214]4+ (with Ka ranging from 44 M-1 to 529 M-1).

Ebselen derivatives inhibit SARS-CoV-2 replication by inhibition of its essential proteins: PLpro and Mpro proteases, and nsp14 guanine N7-methyltransferase.

Proteases encoded by SARS-CoV-2 constitute a promising target for new therapies against COVID-19. SARS-CoV-2 main protease (Mpro, 3CLpro) and papain-like protease (PLpro) are responsible for viral polyprotein cleavage-a process crucial for viral survival and replication. Recently it was shown that 2-phenylbenzisoselenazol-3(2H)-one (ebselen), an organoselenium anti-inflammatory small-molecule drug, is a potent, covalent inhibitor of both the proteases and its potency was evaluated in enzymatic and antiviral assays. In this study, we screened a collection of 34 ebselen and ebselen diselenide derivatives for SARS-CoV-2 PLpro and Mpro inhibitors. Our studies revealed that ebselen derivatives are potent inhibitors of both the proteases. We identified three PLpro and four Mpro inhibitors superior to ebselen. Independently, ebselen was shown to inhibit the N7-methyltransferase activity of SARS-CoV-2 nsp14 protein involved in viral RNA cap modification. Hence, selected compounds were also evaluated as nsp14 inhibitors. In the second part of our work, we employed 11 ebselen analogues-bis(2-carbamoylaryl)phenyl diselenides-in biological assays to evaluate their anti-SARS-CoV-2 activity in Vero E6 cells. We present their antiviral and cytoprotective activity and also low cytotoxicity. Our work shows that ebselen, its derivatives, and diselenide analogues constitute a promising platform for development of new antivirals targeting the SARS-CoV-2 virus.

Convergence of multiple synthetic paradigms in a universally programmable chemical synthesis machine.

Although the automatic synthesis of molecules has been established, each reaction class uses bespoke hardware. This means that the connection of multi-step syntheses in a single machine to run many different protocols and reactions is not possible, as manual intervention is required. Here we show how the Chemputer synthesis robot can be programmed to perform many different reactions, including solid-phase peptide synthesis, iterative cross-coupling and accessing reactive, unstable diazirines in a single, unified system with high yields and purity. Developing universal and modular hardware that can be automated using one software system makes a wide variety of batch chemistry accessible. This is shown by our system, which performed around 8,500 operations while reusing only 22 distinct steps in 10 unique modules, with the code able to access 17 different reactions. We also demonstrate a complex convergent robotic synthesis of a peptide reacted with a diazirine-a process requiring 12 synthetic steps.

Discovering New Chemistry with an Autonomous Robotic Platform Driven by a Reactivity-Seeking Neural Network.

We present a robotic chemical discovery system capable of navigating a chemical space based on a learned general association between molecular structures and reactivity, while incorporating a neural network model that can process data from online analytics and assess reactivity without knowing the identity of the reagents. Working in conjunction with this learned knowledge, our robotic platform is able to autonomously explore a large number of potential reactions and assess the reactivity of mixtures, including unknown chemical spaces, regardless of the identity of the starting materials. Through the system, we identified a range of chemical reactions and products, some of which were well-known, some new but predictable from known pathways, and some unpredictable reactions that yielded new molecules. The validation of the system was done within a budget of 15 inputs combined in 1018 reactions, further analysis of which allowed us to discover not only a new photochemical reaction but also a new reactivity mode for a well-known reagent (p-toluenesulfonylmethyl isocyanide, TosMIC). This involved the reaction of 6 equiv of TosMIC in a "multistep, single-substrate" cascade reaction yielding a trimeric product in high yield (47% unoptimized) with the formation of five new C-C bonds involving sp-sp2 and sp-sp3 carbon centers. An analysis reveals that this transformation is intrinsically unpredictable, demonstrating the possibility of a reactivity-first robotic discovery of unknown reaction methodologies without requiring human input.

Organic synthesis in a modular robotic system driven by a chemical programming language.

The synthesis of complex organic compounds is largely a manual process that is often incompletely documented. To address these shortcomings, we developed an abstraction that maps commonly reported methodological instructions into discrete steps amenable to automation. These unit operations were implemented in a modular robotic platform by using a chemical programming language that formalizes and controls the assembly of the molecules. We validated the concept by directing the automated system to synthesize three pharmaceutical compounds, diphenhydramine hydrochloride, rufinamide, and sildenafil, without any human intervention. Yields and purities of products and intermediates were comparable to or better than those achieved manually. The syntheses are captured as digital code that can be published, versioned, and transferred flexibly between platforms with no modification, thereby greatly enhancing reproducibility and reliable access to complex molecules.

An autonomous organic reaction search engine for chemical reactivity.

The exploration of chemical space for new reactivity, reactions and molecules is limited by the need for separate work-up-separation steps searching for molecules rather than reactivity. Herein we present a system that can autonomously evaluate chemical reactivity within a network of 64 possible reaction combinations and aims for new reactivity, rather than a predefined set of targets. The robotic system combines chemical handling, in-line spectroscopy and real-time feedback and analysis with an algorithm that is able to distinguish and select the most reactive pathways, generating a reaction selection index (RSI) without need for separate work-up or purification steps. This allows the automatic navigation of a chemical network, leading to previously unreported molecules while needing only to do a fraction of the total possible reactions without any prior knowledge of the chemistry. We show the RSI correlates with reactivity and is able to search chemical space using the most reactive pathways.

Synthesis, Structure, and Complexation Properties of a C3-Symmetrical Triptycene-Based Anion Receptor: Selectivity for Dihydrogen Phosphate.

A new anion binding motif based on triptycene core has been synthesized from 2,7,14-trinitrotriptycene. Its well-defined binding pocket allowed for the selective recognition and sensing of dihydrogen phosphate in DMSO-d(6) + 0.5% H(2)O.

Sweet anion receptors: recognition of chiral carboxylate anions by D-glucuronic-acid-decorated diindolylmethane.

Anion receptors containing glucuronic acid were synthesized, and their anion binding ability studied. Chirality of anionic guests derived from mandelic acid and amino acids can be distinguished not only in terms of stability constants but also by significant differences in chemical shift changes for sugar moiety protons.