This repository contains the code and data for the paper: "Contextual evidence for categorising interactions and guiding their discovery".
Abramov, K., Nehoray, S. M., Puzis, R., Mucha, P. J., & Pilosof, S. (2026). Contextual evidence for categorising interactions and guiding their discovery. EcoEvoRxiv. https://doi.org/10.32942/X2JD60
Ecological interactions shape community dynamics and stability. Consequently, our incomplete knowledge of true interaction structure has spurred a surge in link prediction. Yet predictions remain conjectures until validated in the field, and validating them all is neither feasible nor desirable, because links differ in their consequences for community structure and function. Validation therefore requires an operative framework grounded in strong ecological and statistical theory. We present a guided-sampling framework combining link prediction with within-system variability across replicated networks as contextual evidence, generating a fine-resolution, ecologically-aware link taxonomy. It resolves categories ecologists have long struggled to separate (e.g., forbidden versus missing interactions), pinpoints the most probable unknowns, and directs targeted, cost-efficient discovery. We include a Bayesian analysis quantifying the confidence in link categorisation, and enabling the use of researchers' knowledge. We provide a worked empirical analysis and an online guide at http://lpguide.ecomplab.com.
Both ship with a standalone interactive explorer, live at https://lpguide.ecomplab.com/.
Bayesian reading of the link taxonomy. Accompanies the development of Section Bayesian Inference & Supplementary Information. Figure script, rendered figure, the interactive explorer, and a math/assumptions context doc.
The evidence accumulation framework in action. Accompanies Section Empirical Demonstration: the framework applied to plant–pollinator networks at six sites: catrgorisation and prediction scripts, rendered alluvial figure, and the interactive alluvial explorer.
Instructions for running the code and reproducing the results are in the repository Wiki under "Code".
Data used for empirical demonstration is detailed in the repository Wiki under "Data".