Construction of a web-based nanomaterial database by big data curation and modeling friendly nanostructure...

Area of Science:

• Nanotechnology
• Materials Science
• Computational Chemistry

Background:

• Extensive experimental data exist for nanomaterials.
• Current nanomaterial databases lack suitability for computational modeling due to curation methods.

Purpose of the Study:

• To construct a large, annotated nanomaterial database tailored for computational modeling and machine learning.
• To provide a publicly accessible resource for advancing nanoinformatics research.

Main Methods:

• Curated experimental data for 705 unique nanomaterials across 11 material types.
• Annotated nanostructures transformed into downloadable Protein Data Bank (PDB) files.
• Generated 2142 nanodescriptors for machine learning applications.

Main Results:

• Established a comprehensive nanomaterial database with over ten endpoints per nanomaterial, including physicochemical properties and bioactivities.
• Made the database and associated nanodescriptors publicly available via http://www.pubvinas.com/.
• Facilitated global researcher access to annotated nanostructures and data for computational studies.

Conclusions:

• The developed database serves as a crucial public resource for data-driven nanoinformatics.
• Enables rational nanomaterial design and accelerates research in computational nanotechnology.
• Supports the growing field of machine learning applications in materials science.