Accelerated Discovery of the Polymer Blends for Cartilage Repair through Data-Mining Tools and Machine-Learning...

Area of Science:

• Biomaterials Science
• Tissue Engineering
• Computational Biology

Background:

• Scaffold material selection is critical for successful cartilage tissue engineering.
• Empirical methods for identifying suitable polymers are time-consuming and resource-intensive.
• A comprehensive understanding of polymer properties for cartilage repair is needed.

Purpose of the Study:

• To implement an inverse design approach using machine learning to predict optimal polymer scaffolds for cartilage repair.
• To overcome the limitations of traditional empirical selection methods.
• To identify specific polymer(s)/blend(s) suitable for cartilage regeneration.

Main Methods:

• Systematic bibliometric analysis of cartilage repair literature using the bibliometrix R package.
• Database creation by data mining mechanical properties of polymers from the PoLyInfo library.
• Application of a multinomial logistic regression (MNLR) algorithm with polymer mechanical properties as input.

Main Results:

• The study successfully applied an inverse design approach to predict polymer scaffolds.
• The MNLR algorithm identified polyethylene/polyethylene-graftpoly(maleic anhydride) blend as a top candidate.
• Mechanical properties of polymers were correlated with their suitability for cartilage repair.

Conclusions:

• Machine learning, particularly MNLR, offers an efficient inverse design strategy for scaffold material selection.
• Polyethylene/polyethylene-graftpoly(maleic anhydride) shows potential as a biomaterial for cartilage tissue engineering.
• This approach can accelerate the development of effective cartilage repair strategies.