PopGenAgent: An Agentic Approach for Reproducible and Report-Oriented Population Genetics Analysis

Population genetics inference routinely requires the coordination of many specialized tools, the management of heterogeneous and fragile file formats, iterative diagnostic evaluation, and the conversion of intermediate outputs into interpretable figures and written summaries. Although workflow frameworks improve reproducibility, substantial manual effort is still required for parameter selection, troubleshooting, and report preparation. Here, we present PopGenAgent, a practical system for report-oriented population genetics analysis that organizes a curated library of commonly used toolchains into validated execution and visualization templates with standardized input and output contracts and complete provenance capture. PopGenAgent separates retrieval-augmented user assistance for interpretation and report preparation from conservative template-based execution that emphasizes auditable commands, checks of intermediate artefact integrity, and generation of figures suitable for reporting. To reduce computational cost, a smaller language model is used for template selection, parameter instantiation, and limited error correction, whereas higher-capacity models can be invoked selectively for narrative report generation grounded in recorded artefacts. We evaluated PopGenAgent on a broad panel of routine and advanced tasks spanning preprocessing, population structure analysis, and allele-sharing statistics, and further examined its performance through end-to-end replication of standard analyses across 26 populations from the 1000 Genomes Project. The system reproduced expected summaries and patterns, including ROH and heterozygosity profiles, LD decay, PCA, ADMIXTURE structure, TreeMix diagnostics, and $f$-statistics. Together, these results indicate that a validated template library combined with provenance-aware reporting can reduce the manual burden of scripting and workflow coordination while preserving reproducibility and stepwise inspectability in population genetics studies.