S2-PepAnalyst: A Web Tool for Predicting Plant Small Signalling Peptides

Abstract Small signalling peptides mediate cell-to-cell communication playing essential roles in plant growth, development, and stress responses. They specifically bind to the extracellular domain of receptors to trigger biochemical and physiological responses. Despite their significance, accurately identifying novel signalling peptides remains challenging due to their structural diversity, low abundance, and highly specific expression patterns. Here, we present S²-PepAnalyst, a web tool integrating plant-specific datasets and machine learning to predict SSPs with 99.5% accuracy and low false-negative rates. S²-PepAnalyst outperforms existing tools (e.g., SignalP 6.0) by combining protein language models, geometric-topological analysis, and reinforcement learning, enabling robust classification of small signalling peptide families (e.g., CLE, RALF). The tool is freely available at https://www.s2-pepanalyst.uma.es. Competing Interest Statement The authors have declared no competing interest. Footnotes The revised manuscript addresses the web tool's availability, and the optimal protein representation, i.e., combining ESM-2, TAPE, and GeoTop features, is clearly defined. Terminology inconsistencies, such as "success rate", have been resolved with precise definitions, and ambiguous phrases like "non-overfitted feature extraction" have been replaced with clearer explanations. The dataset and model rationale have been expanded to justify the focus on Arabidopsis thaliana and subtropical crops, with plans to include major agricultural species in future updates. Sequences shorter than 10 amino acids were removed, and a new figure illustrates length distributions for SSPs versus non-SSPs. The use of reinforcement learning (RL) is now better justified. Benchmarking has been strengthened with comparisons to methods like SCREW, alongside restructured figures for clearer performance visualisation. Concerns about dataset size and composition were addressed, emphasising the model's robustness across peptide families. The web server's accessibility issues were resolved, and the GitHub repository was reorganised for better usability. Terminology was refined to align with community standards, distinguishing small signalling peptides (SSPs) from non-signalling secreted peptides. The manuscript now acknowledges foundational studies and clarifies how the tool complements SignalP by predicting bioactive SSPs rather than just secretion signals. To bridge the gap between computational and biological audiences, the manuscript was restructured for clarity, with a graphical abstract and a user-friendly web tool guide. Key biological features, such as secretion signals and cysteine-rich motifs, are highlighted, and the tool's ability to predict post-translational modifications (PTMs) like sulfation and disulfide bonds is explained. A new Discussion section addresses limitations, biological implications, and future directions, such as incorporating molecular dynamics simulations. Minor corrections, including typos and additional citations, were made to enhance accuracy. These revisions ensure the manuscript is more accessible, methodologically sound, and relevant to both computational and plant biology researchers.