Rewiring chemistry: algorithmic discovery and experimental validation of one-pot reactions in the network of organic chemistry

Computational algorithms are used to identify sequences of reactions that can be performed in one pot. These predictions are based on over 86?000 chemical criteria by which the putative sequences are evaluated. The "raw" algorithmic output is then validated experimentally by performing multiple two-, three-, and even four-step sequences. These sequences "rewire" synthetic pathways around popular and/or important small molecules.     

Parallel Optimization of Synthetic Pathways within the Network of Organic Chemistry

Finding a needle in a haystack: The number of possible synthetic pathways leading to the desired target of a synthesis can be astronomical (10(19) within five synthetic steps). Algorithms are described that navigate through the entire known chemical-synthetic knowledge to identify optimal synthetic pathways. Examples are provided to illustrate single-target optimization and parallel optimization of syntheses leading to multiple targets.     

Reduced bio basis function neural network for identification of protein phosphorylation sites: comparison with pattern recognition algorithms

Protein phosphorylation is a post-translational modification performed by a group of enzymes known as the protein kinases or phosphotransferases (Enzyme Commission classification 2.7). It is essential to the correct functioning of both proteins and cells, being involved with enzyme control, cell signalling and apoptosis. The major problem when attempting prediction of these sites is the broad substrate specificity of the enzymes. This study employs back-propagation neural networks (BPNNs), the decision tree algorithm C4.5 and the reduced bio-basis function neural network (rBBFNN) to predict phosphorylation sites. The aim is to compare prediction efficiency of the three algorithms for this problem, and examine knowledge extraction capability. All three algorithms are effective for phosphorylation site prediction. Results indicate that rBBFNN is the fastest and most sensitive of the algorithms. BPNN has the highest area under the ROC curve and is therefore the most robust, and C4.5 has the highest prediction accuracy. C4.5 also reveals the amino acid 2 residues upstream from the phosporylation site is important for serine/threonine phosphorylation, whilst the amino acid 3 residues upstream is important for tyrosine phosphorylation.     

Chemical Probes for the Functionalization of Polyketide Intermediates

A library of functionalized chemical probes capable of reacting with ketosynthase‐bound biosynthetic intermediates was prepared and utilized to explore in vivo polyketide diversification. Fermentation of ACP mutants of S. lasaliensis in the presence of the probes generated a range of unnatural polyketide derivatives, including novel putative lasalocid A derivatives characterized by variable aryl ketone moieties and linear polyketide chains (bearing alkyne/azide handles and fluorine) flanking the polyether scaffold. By providing direct information on microorganism tolerance and enzyme processing of unnatural malonyl‐ACP analogues, as well as on the amenability of unnatural polyketides to further structural modifications, the chemical probes constitute invaluable tools for the development of novel mutasynthesis and synthetic biology.     

Maximum common subgraph isomorphism algorithms for the matching of chemical structures

The maximum common subgraph (MCS) problem has become increasingly important in those aspects of chemoinformatics that involve the matching of 2D or 3D chemical structures. This paper provides a classification and a review of the many MCS algorithms, both exact and approximate, that have been described in the literature, and makes recommendations regarding their applicability to typical chemoinformatics tasks.     

Photonic Crystals for Chemical Sensing and Biosensing

This Review covers photonic crystals (PhCs) and their use for sensing mainly chemical and biochemical parameters, with a particular focus on the materials applied. Specific sections are devoted to a) a lead‐in into natural and synthetic photonic nanoarchitectures, b) the various kinds of structures of PhCs, c) reflection and diffraction in PhCs, d) aspects of sensing based on mechanical, thermal, optical, electrical, magnetic, and purely chemical stimuli, e) aspects of biosensing based on biomolecules incorporated into PhCs, and f) current trends and limitations of such sensors.     

FragNet,a Contrastive Learning-Based Transformer Model for Clustering,Interpreting, Visualizing,and Navigating Chemical Space

The question of molecular similarity is core in cheminformatics and is usually assessed via a pairwise comparison based on vectors of properties or molecular fingerprints. We recently exploited variational autoencoders to embed 6M molecules in a chemical space, such that their (Euclidean) distance within the latent space so formed could be assessed within the framework of the entire molecular set. However, the standard objective function used did not seek to manipulate the latent space so as to cluster the molecules based on any perceived similarity. Using a set of some 160,000 molecules of biological relevance, we here bring together three modern elements of deep learning to create a novel and disentangled latent space, viz transformers, contrastive learning, and an embedded autoencoder. The effective dimensionality of the latent space was varied such that clear separation of individual types of molecules could be observed within individual dimensions of the latent space. The capacity of the network was such that many dimensions were not populated at all. As before, we assessed the utility of the representation by comparing clozapine with its near neighbors, and we also did the same for various antibiotics related to flucloxacillin. Transformers, especially when as here coupled with contrastive learning, effectively provide one-shot learning and lead to a successful and disentangled representation of molecular latent spaces that at once uses the entire training set in their construction while allowing “similar” molecules to cluster together in an effective and interpretable way.     

Two novel algorithms for the thermogravimetric assessment of polymer degradation under non-isothermal conditions

Two novel algorithms are presented for processing thermogravimetric (TG) data obtained during the degradation of a polymer in a single step mechanism under non-isothermal conditions. The first algorithm assesses three characteristics computed from the TG profile against a theoretical data set, and identifies likely kinetic models to fit the experimental data. The second algorithm provides an iterative arithmetic method to extract the apparent activation energy, E_a, and Arrhenius A-factor, A, from TG data without simplifying assumptions. The algorithms are validated using model data and applied to data for the non-isothermal degradation of poly(ethylene adipate), poly(lactic acid) (PLA) and a food packaging PLA composite formulation containing kenaf, a natural fibre. The analysis of poly(ethylene adipate) produced E_a = 137 kJ mol⁻¹ and log₁₀A = 8.71 (first-order kinetic model). The kenaf fibre destabilizes PLA, lowering its E_a from 190 kJ mol⁻¹ to 150 kJ mol⁻¹ (contracting volume model).     

A biomimetic principle for the chemical modification of metal surfaces: synthesis of tripodal catecholates as analogues of siderophores and mussel adhesion proteins

By following a biomimetic design principle, tetravalent scaffolds based on an adamantyl and trisalkylmethyl core structure have been synthesized. These scaffolds have been coupled to three catecholamines, thus resembling the characteristic tripodal recognition motif of many natural metal binders, such as mussel adhesion proteins and siderophores, for example, enterobactin. Besides this tripodal recognition element, our scaffolds provide a fourth position for the conjugation of effector molecules. These effectors can be conjugated through biocompatible conjugation techniques to the scaffold and can be used to tailor the properties of different metal surfaces for a range of applications, for example, in implant engineering. Herein, we describe the synthesis of several tripodal metal binders and their immobilization on TiO(2) surfaces by using a simple dip-coating procedure. Furthermore, we demonstrate the conjugation of our surface binders to the dye eosin Y as an effector molecule by peptide coupling. The resulting surfaces have been analyzed by using ellipsometry, time-of-flight secondary ion mass spectrometry, IR spectroscopy, and contact-angle measurements to confirm the specific loading on TiO(2) films and nanoparticles with our trivalent surface binders. As a proof of concept, we have demonstrated the functionalization of TiO(2) nanoparticles with the eosin Y dye.     

Chemical synthesis of Helicobacter pylori lipopolysaccharide partial structures and their selective proinflammatory responses

Helicobacter pylori is a common cause of gastroduodenal inflammatory diseases such as chronic gastritis and peptic ulcers and also an important factor in gastric carcinogenesis. Recent reports have demonstrated that bacterial inflammatory processes, such as stimulation with H. pylori lipopolysaccharide (LPS), initiate atherosclerosis. To establish the structures responsible for the inflammatory response of H. pylori LPS, we synthesized various kinds of lipid A structures (i.e., triacylated lipid A and Kdo‐lipid A compounds), with or without the ethanolamine group at the 1‐phosphate moiety, by a new divergent synthetic route. Stereoselective α‐glycosylation of Kdo N‐phenyltrifluoroacetimidate was achieved by use of microfluidic methods. None of the lipid A and Kdo‐lipid A compounds were a strong inducer of IL‐1β, IL‐6, or IL‐8, suggesting that H. pylori LPS is unable to induce acute inflammation. In fact, the lipid A and Kdo‐lipid A compounds showed antagonistic activity against cytokine induction by E. coli LPS, except for the lipid A compound with the ethanolamine group, which showed very weak agonistic activity. On the other hand, these H. pylori LPS partial structures showed potent IL‐18‐ and IL‐12‐inducing activities. IL‐18 has been shown to correlate with chronic inflammation, so H. pylori LPS might be implicated in the chronic inflammatory responses induced by H. pylori. These results also indicated that H. pylori LPS can modulate the immune response: NF‐κB activation through hTLR4/MD‐2 was suppressed, whereas production of IL‐18 and IL‐12 was promoted.     

遗传算法用于铅的化学形态模拟计算

采用一种全局优方法－遗传算法计算了水体中Ｐｂ＾２＋的化学形态分布，讨论了评从是孙数，初值，群休一中个体数，遗传速率，变异速率等参数选择对计算结果的影响，证明了所建立的模型和彩和的算法的可行性。     

Chemical Preparation of Graphene Materials Results in Extensive Unintentional Doping with Heteroatoms and Metals

Chemical synthesis of graphene relies on the usage of various chemical reagents. The initial synthesis step, in which graphite is oxidized to graphite oxide, is achieved by a combination of chemical oxidants and acids. A subsequent chemical reduction step eliminates/reduces most oxygen functionalities to yield graphene. We demonstrate here that these chemical treatments significantly contaminate graphene with heteroatoms/metals, depending on the procedures followed. Contaminations with heteroatoms (N, B, Cl, S) or metals (Mn, Al) were present at relatively high concentrations (up to 3 at %), with their chemical states dependent on the procedures. Such unintentional contaminations (unwanted doping) during chemical synthesis are rarely anticipated and reported, although the heteroatoms/metals may alter the electronic and catalytic properties of graphene. In fact, the levels of unintentionally introduced contaminants on graphene are often higher than typical levels found on intentionally doped graphene. Our findings are important for scientists applying chemical methods to prepare graphene.     

A method and program for mass quantum chemical calculations of protein—ligand docking complexes

A new fast computational method for mass calculations of docking complexes by the AM1/PM3 semiempirical methods is proposed. The computation time is shortened by at least an order of magnitude compared to alternative schemes of quantum chemical calculations. The root-mean-square deviation of the AM1 calculated energies of formation of complexes from the results obtained by conventional diagonalization procedure is at most 0.4 kcal mol⁻¹. Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 2, pp. 418–420, February, 2008.     

一种判断电化学/化学反应体系热力学自洽性的新方法

介绍了一种判断电化学/化学反应体系热力学自洽性的新方法-“秩法”。结合 实际例子描述了该方法的原理、计算步骤。该方法充分利用矩阵“秩”的概念,由 反应系数矩阵的“秩”直接判断出其中的独立组分数,并且仅当寻找到依赖反应时 才需对有关方程进行求解。因此,“秩法”概念上清晰、直观,且由于可以直接调 用矩阵变换的子程序,程序实现也极为容易。该方法对于电化学、化学及化工反应 过程的模拟将非常有意义。