Search

The next major challenge for artificial intelligence in drug development lies in proving its value in real-world settings. A new technology not only supports the generation of novel chemical entities but also accelerates a range of real-world molecular design tasks.

Sand fly vector control strategies are limited. Here, Cecilio et al. use the bacteria Delftia tsuruhatensis TC1 to reduce the ability of sand flies to become infected with Leishmania parasites and effectively transmit them to mammalian hosts.

Children often show milder COVID-19 symptoms, but the underlying mechanistic insights are still lacking. Here the authors profile both pediatric and adult cohorts of COVID-19 patients in Brazil to find that children exhibit higher viral load but stronger and biased cellular immunity, thereby serving clues for the differential responses in children.

Automation of chemistry research has focused on developing robots to execute jobs. Artificial-intelligence technology has now been used not only to control robots, but also to plan their tasks on the basis of simple human prompts.

One of the most likely substrates for metabolic imaging of response to treatment in cancer is glucose, but until now, using hyperpolarized 13C-labelled glucose has been problematic because of the short lifetime of the hyperpolarization in this molecule. Using [U-¹³C, U-²H]glucose, Tiago Rodrigues et al. now show that they are able to image its glycolytic conversion to lactate in two mouse tumor models in vivo, and that in one model, flux is markedly reduced after treatment with the chemotherapeutic drug etoposide.

Genomic and epidemiologic analyses revealed that a novel reassortant viral lineage of the Oropouche virus, which has been circulating in the western Amazon region of Brazil for about a decade, is associated with the recent human outbreaks between 2022 and 2024.

Changing temperature in striatum warped neural activity in time and categorical time judgments in rats. Similar effects on movement were not observed. Striatal dynamics may thus support discrete decisions and not continuous motor control.

Some strains of Mycobacterium tuberculosis seem to be able to avoid host defense systems. Here the authors stratify patients by severity of tuberculosis and find correlations with the level of IL-1β production by macrophages exposed to these isolates.

Machine learning is a powerful tool for the study and design of molecules. Here the authors comment a recent publication in Nature Communications which highlights the challenges of different molecular representations for data-driven property predictions.

Cyclization provides a general strategy for improving peptide proteolytic stability, cell membrane permeability and target binding affinity. Here the authors develop a cyclopropenone-based proximity-driven chemical linker for the site-specific cyclization of phage-displayed peptides.

This analysis leverages a large-scale literature review, text mining, statistics and machine learning to identify trends, shortcomings and future opportunities in developing and deploying inorganic nanoparticles for cancer diagnosis and therapy.

Natural products provide a rich source of leads for drug discovery. Now, a computational method is available that can be used to identify the macromolecular targets of these compounds. Much like medicinal chemists' reasoning, the software infers target information by comparing the substructures with those of drugs and other natural products with known targets.

A strategy for protecting redox-active ortho-quinones, which show promise as anticancer agents but suffer from redox-cycling behaviour and systemic toxicity, has been developed. The ortho-quinones are derivatized to redox-inactive para-aminobenzyl ketols. Upon amine deprotection, an acid-promoted, self-immolative C–C bond-cleaving 1,6-elimination releases the redox-active hydroquinone. The strategy also enables conjugation to a carrier for targeted delivery of ortho-quinone species.

This study reveals a shallow asthenosphere with high ³He/⁴He ratios (>20 R_A) beneath Patagonia. This primordial composition is linked to the Karoo mantle plume from the early stages of the Gondwana breakup.

Tissue resident macrophages are intricately involved in homeostatic processes but also in tissue repair. Here Lee and colleagues show dermis-resident macrophages are a source of thymic stromal lymphopoietin and CCL24, which act on type 2 innate lymphoid cells and eosinophils respectively, to maintain their M2 properties and promote non-healing cutaneous leishmaniasis.

Genetic exchange has been experimentally demonstrated for Leishmania during sand fly development, indicating a meiotic mechanism. Here the authors show that meiosis-related genes HOP1 and HAP2-2 are essential for Leishmania hybridization in vitro and in sand flies and that their deletion in one or both parents hinders mating competence.

The targeted release of bioactive molecules to diseased tissues has the potential to improve therapeutic efficacy, but not all drugs contain a free functional group that can be easily attached to an antibody. Now, a linker technology has been developed to enable the traceless release of tertiary and heteroaryl amine-containing drugs.

Biochemical and cellular assays are often plagued by false positive readouts elicited by nuisance compounds. A significant proportion of those compounds are aggregators. This Review discusses the basis for colloidal aggregation, experimental methods for detecting aggregates and analyses recent progress in computer-based systems for detecting colloidal aggregation with particular emphasis on machine learning [In the online version of this Review originally published, the graphical abstract image was incorrectly credited to ‘Reven T.C. Wurman / Alamy Stock Photo’ this has now been corrected].

The biological understanding of poor prognosis associated with lymph node metastasis in head and neck cancer (HNC) remains crucial. Here, a proteomic characterisation of 140 multisite samples from a 59-HNC patient cohort and machine learning reveals potential biomarkers and metastasis related signatures.

Spatial control of gene expression allows precise control over biological processes. Here, the authors develop an efficient light-responsive formulation based on upconversion nanoparticles, and demonstrate on-demand genetic manipulation in deep brain tissue.

Natural products are a prime source of innovative molecular fragments and privileged scaffolds for drug discovery and chemical biology. Advanced machine-learning approaches can help analyse and design synthetically accessible, natural-product-derived, compound libraries and provide insight into the high selectivity of such compounds.

Here, the authors present results of the ZiBRA-2 project (https://www.zibra2project.org) which is an arbovirus surveillance project, across the Midwest of Brazil using a mobile genomics laboratory, combined with a genomic surveillance training program that targeted post-graduate students, laboratory technicians, and health practitioners in universities and laboratories.

In this Review, Waksman and colleagues describe the structural and mechanistic details of the six secretion systems (types I–VI) of Gram-negative bacteria, the unique mycobacterial type VII secretion system, the chaperone–usher pathway and the curli biogenesis machinery. They discuss both conserved and divergent properties of these systems and their potential as targets of novel antibacterial compounds.

Borges et al. study the dynamics of SARS-CoV-2 in Portugal during 2020. The authors use phylodynamic modelling to infer patterns of SARS-CoV-2 introduction into Portugal and its subsequent spread, providing a comprehensive genomic epidemiology study of the early SARS-CoV-2 epidemic in Portugal.

Artificial intelligence has recently seen numerous applications in synthetic organic chemistry. Advanced pattern-recognition heuristics may facilitate the access to chemical matter of interest and complement chemical intuition in the near future.

Studies employing machine-learning (ML) tools in the chemical sciences often report their evaluations in a heterogeneous way. The evaluation guidelines provided in this Perspective should enable more rigorous ML reporting.

To evade autophagy-mediated killing when inside liver cells, the Plasmodium berghei protein UIS3 binds to a key regulator of the autophagy programme, the host protein LC3, and inhibits its interaction with downstream effectors.

Bright star \(\nu\) Indi shows elevated levels of alpha-process elements, suggesting great age, and is kinematically heated, probably from the merger of a dwarf galaxy with the Milky Way. Chaplin et al. make a case for \(\nu\) Indi being an accurate indicator of the timing for the Gaia–Enceladus merger.

Jonathas Bittencourt et al. report the discovery of a fossil lizard in South America dated to at least 130 mya, about 20 million years earlier than previously thought. This finding suggests early lizards in South America were much more broadly distributed than extant species in this region.