Search

Machine learning (ML) integrated density functional theory (DFT) calculations have recently been used to accelerate the design and discovery of heterogeneous catalysts such as single atom catalysts (SACs) through the establishment of deep structure–activity relationships.

The raffinate in the extraction process contains plenty of acid and metallic ions, the dispose of which causes the resource loss and environmental pollution. The present study proposed an innovative way to solve this problem by using raffinate as the leaching agent in the copper leaching process. The effects of significant leaching parameters were investigated and optimized using the central composite design (CCD) of response surface methodology (RSM). A maximum leaching efficiency of 96.74 % can be obtained under the 53.22 Kg/t of initial acidity, 2.54 h of leaching time and 5.95 of liquid/solid ratio. In addition, the comparative features of samples before and after leaching process were analyzed by using the X-ray diffraction (XRD), sieving test, scanning electron microscope (SEM) and adsorption-desorption isotherms. We found that the majority of copper components have been removed from copper ores, causing the reduction of size distributions. Specifically, the content of

The access to essential medicines remains a problem in many low-income countries for logistic and expiration limits, among other factors. Enabling flexible replenishment and easier supply chain management by on demand manufacturing from stored starting materials provides a solution to this challenge. Recent developments in computer-aided chemical synthesis planning have benefited from machine learning in different aspects. In this manuscript, we use those techniques to perform a combined analysis of a WHO essential medicines list to identify synthetic routes that minimize chemical inventory that would be required to synthesize the all the active pharmaceutical ingredients. We use a synthesis planning tool to perform retrosynthetic analyses for 99 targets and solve a mixed-integer programming problem to select a combination of pathways that uses the minimal number of chemicals. This work demonstrates the technical feasibility of reducing storage of active pharmaceutical ingredients to a minimal inventory of starting materials.

Background: Breast cancer (BC) is the most common malignancy in women with high heterogeneity. The heterogeneity of cancer cells from different BC subtypes has not been thoroughly characterized and there is still no valid biomarker for predicting the prognosis of BC patients in clinical practice. Methods: Cancer cells were identified by calculating single cell copy number variation using the inferCNV algorithm. SCENIC was utilized to infer gene regulatory networks. CellPhoneDB software was used to analyze the intercellular communications in different cell types. Survival analysis, univariate Cox, least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox analysis were used to construct subtype specific prognostic models. Results: Triple-negative breast cancer (TNBC) has a higher proportion of cancer cells than subtypes of HER2+ BC and luminal BC, and the specifically upregulated genes of the TNBC subtype are associated with antioxidant and chemical stress resistance. Key transcription factors (TFs) of tumor cells for three subtypes varied, and most of the TF-target genes are specifically upregulated in corresponding BC subtypes. The intercellular communications mediated by different receptor–ligand pairs lead to an inflammatory response with different degrees in the three BC subtypes. We establish a prognostic model containing 10 genes (risk genes: ATP6AP1, RNF139, BASP1, ESR1 and TSKU; protective genes: RPL31, PAK1, STARD10, TFPI2 and SIAH2) for luminal BC, seven genes (risk genes: ACTR6 and C2orf76; protective genes: DIO2, DCXR, NDUFA8, SULT1A2 and AQP3) for HER2+ BC, and seven genes (risk genes: HPGD, CDC42 and PGK1; protective genes: SMYD3, LMO4, FABP7 and PRKRA) for TNBC. Three prognostic models can distinguish high-risk patients from low-risk patients and accurately predict patient prognosis. Conclusions: Comparative analysis of the three BC subtypes based on cancer cell heterogeneity in this study will be of great clinical significance for the diagnosis, prognosis and targeted therapy for BC patients.

Access to structured chemical reaction data is of key importance for chemists in performing bench experiments and in modern applications like computer-aided drug design. Existing reaction databases are generally populated by human curators through manual abstraction from published literature (e.g., patents and journals), which is time consuming and labor intensive, especially with the exponential growth of chemical literature in recent years. In this study, we focus on developing automated methods for extracting reactions from chemical literature. We consider journal publications as the target source of information, which are more comprehensive and better represent the latest developments in chemistry compared to patents; however, they are less formulaic in their descriptions of reactions. To implement the reaction extraction system, we first devised a chemical reaction schema, primarily including a central product, and a set of associated reaction roles such as reactants, catalyst, solvent, and so on. We formulate the task as a structure prediction problem and solve it with a two-stage deep learning framework consisting of product extraction and reaction role labeling. Both models are built upon Transformer-based encoders, which are adaptively pretrained using domain and task-relevant unlabeled data. Our models are shown to be both effective and data efficient, achieving an F1 score of 76.2% in product extraction and 78.7% in role extraction, with only hundreds of annotated reactions.

Computer aided synthesis planning of synthetic pathways with green process conditions has become of increasing importance in organic chemistry, but the large search space inherent in synthesis planning and the difficulty in predicting reaction conditions make it a significant challenge. We introduce a new Monte Carlo Tree Search (MCTS) variant that promotes balance between exploration and exploitation across the synthesis space. Together with a value network trained from reinforcement learning and a solvent-prediction neural network, our algorithm is comparable to the best MCTS variant (PUCT, similar to Google's Alpha Go) in finding valid synthesis pathways within a fixed searching time, and superior in identifying shorter routes with greener solvents under the same search conditions. In addition, with the same root compound visit count, our algorithm outperforms the PUCT MCTS by 16% in terms of determining successful routes. Overall the success rate is improved by 19.7% compared to the upper confidence bound applied to trees (UCT) MCTS method. Moreover, we improve 71.4% of the routes proposed by the PUCT MCTS variant in pathway length and choices of green solvents. The approach generally enables including Green Chemistry considerations in computer aided synthesis planning with potential applications in process development for fine chemicals or pharmaceuticals., A new MCTS variant with a reinforcement learning value network and solvent prediction model proposes shorter synthesis routes with greener solvents.

Obesity is caused by an imbalance between food intake and energy expenditure (EE). Here we identify a conserved pathway that links signalling through peripheral Y1 receptors (Y1R) to the control of EE. Selective antagonism of peripheral Y1R, via the non-brain penetrable antagonist BIBO3304, leads to a significant reduction in body weight gain due to enhanced EE thereby reducing fat mass. Specifically thermogenesis in brown adipose tissue (BAT) due to elevated UCP1 is enhanced accompanied by extensive browning of white adipose tissue both in mice and humans. Importantly, selective ablation of Y1R from adipocytes protects against diet-induced obesity. Furthermore, peripheral specific Y1R antagonism also improves glucose homeostasis mainly driven by dynamic changes in Akt activity in BAT. Together, these data suggest that selective peripheral only Y1R antagonism via BIBO3304, or a functional analogue, could be developed as a safer and more effective treatment option to mitigate diet-induced obesity. Neuropeptide Y signalling in the periphery contributes to the regulation of metabolic and energy homeostasis. Here the authors show that blocking Y1R signalling in peripheral tissues using the selective antagonist BIBO3304 ameliorates diet-induced obesity and improves whole-body glucose metabolism.

The synthesis of thousands of candidate compounds in drug discovery and development offers opportunities for computer-aided synthesis planning to simplify the synthesis of molecule libraries by leveraging common starting materials and reaction conditions. We develop an optimization-based method to analyze large organic chemical reaction networks and design overlapping synthesis plans for entire molecule libraries so as to minimize the overall number of unique chemical compounds needed as either starting materials or reaction conditions. We consider multiple objectives, including the number of starting materials, the number of catalysts/solvents/reagents, and the likelihood of success of the overall syntheses plan, to select an optimal reaction network to access the target molecules. The library synthesis planning task was formulated as a network flow optimization problem, and we design an efficient decomposition scheme that reduces solution time by a factor of 5 and scales to instance with 48 target molecules and nearly 8000 intermediate reactions within hours. In four case studies of pharmaceutical compounds, the approach reduces the number of starting materials and catalysts/solvents/reagents needed by 32.2 and 66.0% on average and up to 63.2 and 80.0% in the best cases. The code implementation can be found at https://github.com/Coughy1991/Molecule_library_synthesis.

Computer-aided synthesis has received much attention in recent years. It is a challenging topic in itself, due to the high dimensionality of chemical and reaction space. It becomes even more challenging when the aim is to suggest syntheses that can be performed in continuous flow. Though continuous flow offers many potential benefits, not all reactions are suited to be operated continuously. In this work, three machine learning models have been developed to provide an assessment of whether a given reaction may benefit from continuous operation, what the likelihood of success in continuous flow is for a certain set of reaction components (i.e. reactants, reagents, solvents, catalysts, and products) and, if the likelihood of success is low, which alternative reaction components can be considered. The first model uses an abstract version of a reaction template, obtained via gaussian mixture modelling, to quantify its relative increase in publishing frequency in continuous flow, without relying on potentially ambiguously defined reaction templates. The second model is an artificial neural network that categorizes feasible and infeasible reaction components with a 75 % success rate. A set of reaction components is considered to be feasible if there is an explicit reference to it being used in continuous synthesis in the database; all other reaction components are considered infeasible. While several cases that are ‘infeasible’ by this definition, are classified as feasible by the neural network, further analysis shows that for many of these cases, it is at least plausible that they are in fact feasible – they simply have not been tested to (dis)prove this. The final model suggests alternative continuous flow components with a top-1 accuracy of 95%. Combined, they offer a black-box evaluation of whether a reaction and a set of reaction components can be considered promising for continuous syntheses.

Reaction condition recommendation is an essential element for the realization of computer-assisted synthetic planning. Accurate suggestions of reaction conditions are required for experimental validation and can have a significant effect on the success or failure of an attempted transformation. However, de novo condition recommendation remains a challenging and under-explored problem and relies heavily on chemists’ knowledge and experience. In this work, we develop a neural-network model to predict the chemical context (catalyst(s), solvent(s), reagent(s)), as well as the temperature most suitable for any particular organic reaction. Trained on ∼10 million examples from Reaxys, the model is able to propose conditions where a close match to the recorded catalyst, solvent, and reagent is found within the top-10 predictions 69.6% of the time, with top-10 accuracies for individual species reaching 80–90%. Temperature is accurately predicted within ±20 °C from the recorded temperature in 60–70% of test cases, with higher accuracy for cases with correct chemical context predictions. The utility of the model is illustrated through several examples spanning a range of common reaction classes. We also demonstrate that the model implicitly learns a continuous numerical embedding of solvent and reagent species that captures their functional similarity., Machine learning model predicts conditions for organic synthesis reactions and quantifies solvent and reagent similarity.

A moving bed biofilm reactor (MBBR) using Mn(II) and Fe(II) as mixed electron donors was designed for nitrate removal. The optimal state, as determined by response surface methodology, was an Fe(II):Mn(II) molar ratio of 0.62, electron donor:electron acceptor molar ratio of 2.62 and hydraulic retention time of 10.88 h. Subsequently, the MBBR was applied to groundwater treatment and demonstrated a final nitrate-N removal efficiency of 99.5% with a nitrite-N accumulation rate of 0.0706 mg-N·L−1·h−1. Furthermore, high-throughput sequencing was employed to characterize bacterial communities in the MBBR. Results showed that the genera of Pseudomonas and Acinetobacter may make a contribution to the nitrate removal.

The ions activation of quartz is a thorny problem that deteriorates the flotation of metal ore. Polymer inhibitors have exhibited excellent depression performance in the selective flotation of metal-ion activated minerals. In this study, the inhibitory role of sodium alginate (NaAl) in the flotation of Ca2+-activated quartz was thoroughly explored by zeta potential, dynamic light scattering (DLS), Fourier transform infrared (FTIR) analysis, X-ray photoelectron spectroscopy (XPS) analysis and adsorption test. The flotation results suggested that NaAl could inhibit the flotation of Ca2+-activated quartz in sodium oleate (NaOl) system to different degrees, depending on the addition sequence of NaOl and NaAl. More significant depression of Ca2+-activated quartz could be achieved when NaAl was added prior to NaOl. It was found that prominent mutual interaction existed among Ca2+, NaOl and NaAl in the bulk solution, probably forming calcium alginate–calcium oleate crosslink product, which was not affected by the addition sequence of these two reagents. However, the interfacial behaviors of these two reagents on Ca2+-activated quartz surfaces differed under different reagent conditions. The addition of NaAl after NaOl could not only partially desorb the pre-adsorbed NaOl from mineral surfaces, but also decrease the mineral surface hydrophobicity by polymer coating, which therefore deteriorated the flotation of Ca2+-activated quartz. In contrast, when NaAl was added prior to NaOl, the pre-adsorption of NaAl made the further adsorption of NaOl rather difficult, which was likely attributed to the reticulated structure of NaAl and the associated steric hindrance effect. As a result, the flotation of Ca2+-activated quartz was tightly depressed.

The diversity of the potential arrangements of multiple monomers along the length of polymer chains and their impact on polymer properties spark interest in the design of polymer sequence characteristics for particular applications. Kinetic Monte Carlo (KMC) is a technique that can track the explicit arrangement of monomers in the polymer chains, yet it is difficult to integrate with conventional gradient-based optimization algorithms that are typically invoked to design polymer properties. In this work, we applied and compared derivative-free optimization algorithms to incorporate KMC simulations and find synthesis conditions for achieving property targets and minimizing reaction time, advancing our ability to carry out the design of polymer microstructures and control polymerization processes.

The modeling of the molecular weight distribution of polymers is critical in polymerization modeling. Moment-based models and kinetic Monte Carlo are two methods used to calculate the number and weight average molecular weight during simulation of polymerization. While these two methods are commonly considered to yield equivalent results in the molecular weight distribution, we show that there can be significant differences between these two methods when copolymerization of multiple monomers is modeled. In particular, clear differences in results exist for low molecular weight polymers where the molar mass of one monomer is two or more times greater than the others. Detailed analysis is provided for the cause of the difference, and we demonstrate the importance of tracking the exact number of each type of monomer in every polymer chain when calculating the moments of molecular weight.