Explainable AI-Guided Bioink Design: Model-Derived Polymer-Pressure Relationships in Alginate-Hyaluronic Acid Bioinks

Abstract

Explainable AI-Guided Bioink Design Initial journal reproducibility release accompanying the manuscript: Explainable AI-Guided Bioink Design: Model-Derived Polymer-Pressure Relationships in Alginate-Hyaluronic Acid Bioinks Included materials Curated rheology, two-dimensional printability, and angular-fidelity datasets Processed target-specific modeling datasets for Pr, SR, Qm, and AF Feature-engineering definitions and feature dictionaries Repeated five-fold cross-validation results Leave-one-formulation-out validation results Model-selection and full-fit prediction outputs SHAP and permutation-importance analyses XAI-derived design-hypothesis summaries Candidate-screening and strict-candidate outputs Engineered-feature ablation analyses Analysis notebooks, Python scripts, supplementary tables, and figure-generation materials Reproducibility documentation, environment files, and file manifest Scientific scope This repository supports a secondary, target-specific machine-learning and explainable-AI reanalysis of the alginate-hyaluronic acid bioink dataset reported by Perin et al. No new experimental measurements were generated in the present study. Candidate rankings are model-derived, hypothesis-generating priorities within the experimentally supported source-data domain and should not be interpreted as experimentally confirmed optima. Source data Perin, F.; Spessot, E.; Famà, A.; Bucciarelli, A.; Callone, E.; Mota, C.; Motta, A.; Maniglio, D. Modeling a Dynamic Printability Window on Polysaccharide Blend Inks for Extrusion Bioprinting. ACS Biomaterials Science & Engineering 2023, 9, 1320-1331. DOI: 10.1021/acsbiomaterials.2c01143. Publisher PDFs are not redistributed in this repository. Version Version: v1.0.1Release date: September 18, 2026