EU Cosmetics Regulation establishing the ban on animal testing for cosmetics ingredients requires non-testing (in vitro and in silico) safety assessment procedures. Thus, the “Safety Evaluation Ultimately Replacing Animal Testing” (SEURAT-1) cluster was funded by the EU 7<sup>th</sup> FP and Cosmetics Europe. The COSMOS project (2011-2016) was one of the cluster consortia initiated to facilitate the prediction of human repeated dose toxicity after exposure to cosmetics-related compounds via in silico approaches. Its major outcome, the COSMOS DB (continuously maintained via COSMOS DataShare Point), provides management and sharing of chemical, biological and toxicity data to support this objective. The chemistry content of the COSMOS DB (81,604 chemical records) includes the cosmetics inventories compiled from US Personal Care Products Council (PCPC) and Cosmetics Ingredient Review (CIR), EU CosIng, and Korean Cosmetics Industry Institute (KCII). The toxicity part (including in vivo oral repeated dose toxicity database, oRepeatToxDB) contains dose-level effects information collected, curated, quality-controlled, stored and managed in a flexible and sustainable manner to support predictive modeling tasks. It is well suited for data mining to elucidate site/effect combinations and to be used as a basis for building toxicity ontology for mechanistically relating the chemistry to effects in toxicity pathways. In this study, the chemical and biological activity space based on phys/chem properties and ToxPrint chemotypes was characterized for the cosmetics ingredients inventories (over 7,000 unique structures) with PCA and 2D clustering to identify overlaps and areas of unique features between the inventories. The data-rich regulatory inventories (food-related chemicals from EU EFSA, US FDA’s PAFA, EU ECHA REACH, in addition to the toxicity data from COSMOS DB) were projected onto this space to determine toxicity data-rich regions. The cosmetics inventories showed significant overlap, but only 10% of the structures were common across all of them. The geographical diversity of chemical space can be thus leveraged to expand the general data profile. The repeated-dose toxicity data-rich chemical space of the inventories has been determined for potential increase of confidence during in silico modeling and read-across activities by increasing of the number and quality of available data points.

COSMOS database as a tool for ontology-driven data mining, in silico modeling and read-across

254th National Meeting (2017)

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