Worldwide, hair-coloration represents a multi-billion-dollar industry. Synthetic hair dyes are typically classified according to their shampoo resistance (temporary, semi-permanent, or permanent) and/or their chemical properties (non-oxidative or oxidative). Approximately, 80% of hair dyes are permanent, and many of their small molecule precursors are known skin-sensitizers that can cause severe allergic reactions in certain consumers. There is a global need for new permanent hair dyes of low cost and toxicity, and high efficacy. Herein, our research aims at developing the next generation of hair dyes. First, we developed the Hair Dye Substance Database (HDSD) as a freely-accessible database of hair dyes with their known experimental properties extracted from publicly-available sources. This cheminformatics-curated database includes 313 compounds built with substances classified as a precursor or dye. Dyes were further classified by type (temporary, semi-permanent, or permanent), color, and chromophore (ACS. Sustainable Chem. Eng. 2018 DOI: 10.1021/acssuschemeng.7b03795). Second, a series of 117 2D RDKit descriptors were calculated and used to cluster substances based on their chemical similarity. The clustering analysis identified two semi-permanent dyes (C.I. Basic Orange 1 and Orange 2) with high similarity to several precursors. Third, we conducted a 2D-chemical similarity search using C.I. Basic Orange 1 and Orange 2 as reference compounds to mine NC State’s Max Weaver Dye Library (MWDL); a library housing over 100K unique physical dyes (Chem. Sci. 2017, DOI: 10.1039/C7SC00567A). This first similarity search, on ~2,700 digitized dyes, identified 158 MWDL compounds and many contained a novel phenylazopyrimidine hair dye scaffolding. Fourth, a secondary search (using HDSD precursors as reference compounds) narrowed the top-MWDL compounds to four dyes. These compounds underwent further computational analyses (e.g., predicting skin sensitization potential, generating electrostatic surface potentials, HOMO-LUMO energies, and electron density) and experimental dye uptake on hair fibers analysis (ACS. Sustainable Chem. Eng. 2018 DOI: 10.1021/acssuschemeng.8b02882). Overall, we demonstrate that cheminformatics methods and tools can be utilized to screen for and rationally design the next generation of hair dyes with optimized properties.

Designing the next generation of hair dyes using cheminformatics

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.