We provided one of the largest datasets with purely experimental log kp and developed reliable and accurate prediction models for screening active ingredients and seeking unsynthesized compounds of dermatological medicines and cosmetics.

Our high-performance prediction model offers an attractive alternative to permeability experiments for pharmaceutical and cosmetic candidate screening and optimizing skin-permeable topical formulations.

This study aimed to determine the potentially severe chemical properties of drugs that can cause adverse drug reactions (ADRs) such as erythema multiforme (EM), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) by using a data mining method. The study data were extracted from the Adverse Event Reporting System database of the FDA. EM was considered a mild reaction, and SJS and TEN were considered severe reactions. In this study, a new concept termed the "risk of aggravation" (ROA) was defined as whether a certain drug is more likely to cause severe adverse reactions than mild ones. Partial least squares and logistic regression analysis were applied using binary response variable ROAs. These analyses correctly predicted 50 of the 72 drugs associated with SJS and/or TEN and 28 of the 38 drugs associated with EM using binary chemical descriptors that are the same as those using the metric chemical descriptors.

No abstract