Search

Abstract Background The increasing amount of chemical reaction data makes traditional ways to navigate its corpus less effective, while the demand for novel approaches and instruments is rising. Recent data science and machine learning techniques support the development of new ways to extract value from the available reaction data. On the one side, Computer-Aided Synthesis Planning tools can predict synthetic routes in a model-driven approach; on the other side, experimental routes can be extracted from the Network of Organic Chemistry, in which reaction data are linked in a network. In this context, the need to combine, compare and analyze synthetic routes generated by different sources arises naturally. Results Here we present LinChemIn, a python toolkit that allows chemoinformatics operations on synthetic routes and reaction networks. Wrapping some third-party packages for handling graph arithmetic and chemoinformatics and implementing new data models and functionalities, LinChemIn allows the interconversion between data formats and data models and enables route-level analysis and operations, including route comparison and descriptors calculation. Object-Oriented Design principles inspire the software architecture, and the modules are structured to maximize code reusability and support code testing and refactoring. The code structure should facilitate external contributions, thus encouraging open and collaborative software development. Conclusions The current version of LinChemIn allows users to combine synthetic routes generated from various tools and analyze them, and constitutes an open and extensible framework capable of incorporating contributions from the community and fostering scientific discussion. Our roadmap envisages the development of sophisticated metrics for routes evaluation, a multi-parameter scoring system, and the implementation of an entire “ecosystem” of functionalities operating on synthetic routes. LinChemIn is freely available at https://github.com/syngenta/linchemin. Graphical Abstract

Abstract Patents show how technology evolves in most scientific fields over time. The best way to use this valuable knowledge base is to use efficient and effective information retrieval and searches for related prior art. Patent classification, that is, assigning a patent to one or more predefined categories, is a fundamental step towards synthesizing the information content of an invention. To this end, architectures based on Transformers, especially those derived from the BERT family have already been proposed in the literature and they have shown remarkable results by setting a new state‐of‐the‐art performance for the classification task. Here, we study how domain adaptation can push the performance boundaries in patent classification by rigorously evaluating and implementing a collection of recent transfer learning techniques, for example, domain‐adaptive pretraining and adapters. Our analysis shows how leveraging these advancements enables the development of state‐of‐the‐art models with increased precision, recall, and F1‐score. We base our evaluation on both standard patent classification datasets derived from patent offices‐defined code hierarchies and more practical real‐world use‐case scenarios containing labels from the agrochemical industrial domain. The application of these domain adapted techniques to patent classification in a multilingual setting is also examined and evaluated.

Injectisomes are multi-protein transmembrane machines allowing pathogenic bacteria to inject effector proteins into eukaryotic host cells, a process called type III secretion. Here we present the first three-dimensional structure of Yersinia enterocolitica and Shigella flexneri injectisomes in situ and the first structural analysis of the Yersinia injectisome. Unexpectedly, basal bodies of injectisomes inside the bacterial cells showed length variations of 20%. The in situ structures of the Y. enterocolitica and S. flexneri injectisomes had similar dimensions and were significantly longer than the isolated structures of related injectisomes. The crystal structure of the inner membrane injectisome component YscD appeared elongated compared to a homologous protein, and molecular dynamics simulations documented its elongation elasticity. The ring-shaped secretin YscC at the outer membrane was stretched by 30–40% in situ, compared to its isolated liposome-embedded conformation. We suggest that elasticity is critical for some two-membrane spanning protein complexes to cope with variations in the intermembrane distance.

Transcription Activator-Like (TAL) effectors are DNA-binding proteins secreted by phytopathogenic bacteria that interfere with native cellular functions by binding to plant DNA promoters. The key element of their architecture is a domain of tandem-repeats with almost identical sequences. Most of the polymorphism is located at two consecutive amino acids termed Repeat Variable Diresidue (RVD). The discovery of a direct link between the RVD composition and the targeted nucleotide allowed the design of TAL-derived DNA-binding tools with programmable specificities that revolutionized the field of genome engineering. Despite structural data, the molecular origins of this specificity as well as the recognition mechanism have remained unclear. Molecular simulations of the recent crystal structures suggest that most of the protein-DNA binding energy originates from non-specific interactions between the DNA backbone and non-variable residues, while RVDs contributions are negligible. Based on dynamical and energetic considerations we postulate that, while the first RVD residue promotes helix breaks--allowing folding of TAL as a DNA-wrapping super-helix--the second provides specificity through a negative discrimination of matches. Furthermore, we propose a simple pharmacophore-like model for the rationalization of RVD-DNA interactions and the interpretation of experimental findings concerning shared affinities and binding efficiencies. The explanatory paradigm presented herein provides a better comprehension of this elegant architecture and we hope will allow for improved designs of TAL-derived biotechnological tools.

Capnocytophaga canimorsus is a usual member of dog's mouths flora that causes rare but dramatic human infections after dog bites. We determined the structure of C. canimorsus lipid A. The main features are that it is penta-acylated and composed of a "hybrid backbone" lacking the 4' phosphate and having a 1 phosphoethanolamine (P-Etn) at 2-amino-2-deoxy-d-glucose (GlcN). C. canimorsus LPS was 100 fold less endotoxic than Escherichia coli LPS. Surprisingly, C. canimorsus lipid A was 20,000 fold less endotoxic than the C. canimorsus lipid A-core. This represents the first example in which the core-oligosaccharide dramatically increases endotoxicity of a low endotoxic lipid A. The binding to human myeloid differentiation factor 2 (MD-2) was dramatically increased upon presence of the LPS core on the lipid A, explaining the difference in endotoxicity. Interaction of MD-2, cluster of differentiation antigen 14 (CD14) or LPS-binding protein (LBP) with the negative charge in the 3-deoxy-D-manno-oct-2-ulosonic acid (Kdo) of the core might be needed to form the MD-2 - lipid A complex in case the 4' phosphate is not present.

The enzyme proline racemase from the eukaryotic parasite Trypanosoma cruzi(responsible for endemic Chagas disease) catalyzes the reversible stereoinversion of chiral Cin proline. We employed a new combined quantum mechanical and molecular mechanical (QM/MM) potential to study the reaction mechanism of the enzyme. Three critical points were found:  two almost isoenergetic minima (M1aand M2a), in which the enzyme is bound to l- and d-Pro, respectively, and a transition state (TSCa), unveiling a highly asynchronous concerted process. A systematic analysis was performed on the optimized geometries to point out the key role played by some residues in stabilizing the transition state. [ABSTRACT FROM AUTHOR]