Alasdair D. Wilson

The volume and complexity of scientific literature are expanding rapidly, making it increasingly difficult to extract and synthesise information across studies. This challenge is particularly acute in the biological sciences, where evidence spans multiple levels of organisation and heterogeneous experimental designs. Large Language Model (LLM) pipelines offer a scalable route to evidence synthesis, but many existing approaches lack transparency, modularity, and effective mechanisms for human oversight. We present MetaBeeAI, an open-source, modular pipeline that integrates established LLM techniques into a coherent, auditable workflow for structured data extraction in biology. MetaBeeAI combines modular prompting, multi-pass extraction, and expert-in-the-loop validation within an interface that presents model outputs alongside source text, enabling inspection, correction, and iterative refinement. The pipeline produces machine-readable records of prompts, configurations, and expert annotations, supporting reproducibility and continuous improvement. We apply MetaBeeAI to 924 research papers on bees and pesticides, extracting structured information on species, compounds, exposure designs, and experimental context. Evaluation demonstrates improved consistency, convergence with expert judgement, and robustness across heterogeneous biological studies, highlighting the value of expert-guided refinement. MetaBeeAI provides a transparent and extensible framework for scalable evidence synthesis, supporting reliable integration of LLMs into biological research workflows.

The observation of radio, X-ray, and H _α emission from substellar objects indicates the presence of plasma regions and associated high-energy processes in their surrounding envelopes. This paper numerically simulates and characterizes critical velocity ionization (CVI), a potential ionization process, that can efficiently generate plasma as a result of neutral gas flows interacting with seed magnetized plasmas. By coupling a gas–magnetohydrodynamic (MHD) interactions code (to simulate the ionization mechanism) with a substellar global circulation model (to provide the required gas flows), we quantify the spatial extent of the resulting plasma regions, their degree of ionization, and their lifetime for a typical substellar atmosphere. It is found that the typical average ionization fraction reached at equilibrium (where the ionization and recombination rates are equal and opposite) ranges from 10⁻⁵ to 10⁻⁸, at pressures between 10⁻¹ and 10⁻³ bar, with a trend of increasing ionization fraction with decreasing atmospheric pressure. The ionization fractions reached as a result of CVI are sufficient to allow magnetic fields to couple to gas flows in the atmosphere.

Partially ionized plasmas are ubiquitous in both nature and the laboratory, and their behaviour is best described by models which take into account the interactions between the neutral and charged species. We present a new non-linear, 3-dimensional, finite difference Gas-MHD Interactions Code designed to solve simultaneously the time evolution of fluid equations of both species in the conservation form as well as collisional interactions between them via appropriate choices of source term; in particular, we present results from this code in simulating Alfvén ionization in a partially ionized plasma. In this fashion, larger changes in the ionization fraction than were addressable in the linear limit are possible. Alfvén ionization is shown to impart plasmas with an inherent resistance to rapid recombination, where the recombination itself is significant enough to drive relative motion between the ionised and neutral species at speeds in excess of the critical velocity.

A numerical model of partially ionized plasmas is developed in order to capture their evolving ionization fractions as a result of Alfvén ionization (AI). The mechanism of, and the parameter regime necessary for, AI is discussed and an expression for the AI rate based on fluid parameters, from a gas-MHD model, is derived. This AI term is added to an existing MHD-gas interactions' code, and the result is a linear, 2D, two-fluid model that includes momentum transfer between charged and neutral species as well as an ionization rate that depends on the velocity fields of both fluids. The dynamics of waves propagating through such a partially ionized plasma are investigated, and it is found that AI has a significant influence on the fluid dynamics as well as both the local and global ionization fraction.