Search for: All records

Pairwise host–parasite relationships are typically embedded in broader networks of ecological interactions, which have the potential to shape parasite evolutionary trajectories. Understanding this ‘community context’ of pathogen evolution is vital for wildlife, agricultural and human systems alike, as pathogens typically infect more than one host—and these hosts may have independent ecological relationships. Here, we introduce an eco-evolutionary model examining ecological feedback across a range of host–host interactions. Specifically, we analyse a model of the evolution of virulence of a parasite infecting two hosts exhibiting competitive, mutualistic or exploitative relationships. We first find that parasite specialism is necessary for inter-host interactions to impact parasite evolution. Furthermore, we find generally that increasing competition between hosts leads to higher shared parasite virulence while increasing mutualism leads to lower virulence. In exploitative host–host interactions, the particular form of parasite specialization is critical—for instance, specialization in terms of onward transmission, host tolerance or intra-host pathogen growth rate lead to distinct evolutionary outcomes under the same host–host interactions. Our work provides testable hypotheses for multi-host disease systems, predicts how changing interaction networks may impact virulence evolution and broadly demonstrates the importance of looking beyond pairwise relationships to understand evolution in realistic community contexts. 

Abstract Achieving large-scale, transformative climate change adaptations in agriculture while mitigating further climate impacts and supporting sustainable and equitable rural livelihoods is a grand challenge for society. Transformation of the agri-food system is necessary and inevitable, but the extent to which transformation can be intentionally guided toward desirable states remains unclear. We argue that, instead of targeting leverage points (LPs) in isolation, coordinated interventions at multiple LPs and their interactions are necessary to create a broader system transformation toward more adaptive futures. Using the southeastern United States of America as a case study, we conceptualize a way of doing transformation research in agri-food systems that integrates multiple theoretical and practical perspectives of how transformative pathways can be constructed from ‘chains’ of interacting LPs. We outline several principles for transformative research, the core of which are participatory, transdisciplinary, and convergence research methods needed for articulating a shared vision. These principles embrace an action-oriented approach to research in which the act of assembling diverse networks of researchers, stakeholders, and community partners itself can activate community- and regional-level LPs to scale up changes. Finally, we present tangible examples of specific LPs and their interactions targeted by agri-food system interventions currently underway or planned. This work offers an ‘anticipatory’ vision for agri-food system transformation research that recognizes the need to normatively create an enabling environment to build momentum toward shared visions of secure, equitable, and sustainable regional agri-food systems. 

Abstract Large language models (LLMs) have shown strong performance in tasks across domains but struggle with chemistry-related problems. These models also lack access to external knowledge sources, limiting their usefulness in scientific applications. We introduce ChemCrow, an LLM chemistry agent designed to accomplish tasks across organic synthesis, drug discovery and materials design. By integrating 18 expert-designed tools and using GPT-4 as the LLM, ChemCrow augments the LLM performance in chemistry, and new capabilities emerge. Our agent autonomously planned and executed the syntheses of an insect repellent and three organocatalysts and guided the discovery of a novel chromophore. Our evaluation, including both LLM and expert assessments, demonstrates ChemCrow’s effectiveness in automating a diverse set of chemical tasks. Our work not only aids expert chemists and lowers barriers for non-experts but also fosters scientific advancement by bridging the gap between experimental and computational chemistry. 

We derive a formulation of molecular dynamics that generates only symmetric configurations. We implement it for all 2D planar and 3D space groups. An atlas of 2D Lennard-Jones crystals under all planar groups is created with symmetric molecular dynamics. 

Co-assembly of oppositely charged peptides (CoOP) links intermolecular interactions and material properties. 

In this work, we investigate the question: do code-generating large language models know chemistry? Our results indicate, mostly yes. To evaluate this, we introduce an expandable framework for evaluating chemistry knowledge in these models, through prompting models to solve chemistry problems posed as coding tasks. To do so, we produce a benchmark set of problems, and evaluate these models based on correctness of code by automated testing and evaluation by experts. We find that recent LLMs are able to write correct code across a variety of topics in chemistry and their accuracy can be increased by 30 percentage points via prompt engineering strategies, like putting copyright notices at the top of files. Our dataset and evaluation tools are open source which can be contributed to or built upon by future researchers, and will serve as a community resource for evaluating the performance of new models as they emerge. We also describe some good practices for employing LLMs in chemistry. The general success of these models demonstrates that their impact on chemistry teaching and research is poised to be enormous.