PhyloSophR
PhyloSophR is a doctoral research project investigating how AI can help scientists formulate hypotheses and design transparent, reproducible studies in phylogenetic comparative biology.
The development plan starts with a defined scope and structured schema, followed by a pilot corpus and annotation protocol. A benchmark comes before a sophisticated engine: candidate approaches will be compared with simpler baselines before decisions about retrieval or fine-tuning.
The next planned stages are a minimum engine, human-in-the-loop review, and then simulation and power analysis. Evaluation will address scientific accuracy, unsupported claims, uncertainty, data leakage, and usability.
The long-term goal is to connect study-design decisions to reproducible analysis and preregistration. Automated workflow export and simulation are intended capabilities, not completed features.
Status: in development. The corpus, benchmark, and engine are research milestones; community annotation is not yet open.