Prediction-Inspired Intelligent Training for the Development of Classification Read-across Structure–Activity Relationship (c-RASAR) Models for Organic Skin Sensitizers: Assessment of Classification Error Rate from Novel Similarity Coefficients

Abstract

The advancements in the field of cheminformatics have led to a reduction in animal testing to estimate the activity, property, and toxicity of query chemicals. Read-across structure− activity relationship (RASAR) is an emerging concept that utilizes various similarity functions derived from chemical information to develop highly predictive models. Unlike quantitative structure− activity relationship (QSAR) models, RASAR descriptors of a query compound are computed from its close congeners instead of the compound itself, thus targeting predictions in the model training phase. The objective of the present study is not to propose new QSAR models for skin sensitization but to demonstrate the enhancement in the quality of predictions of the skin-sensitizing potential of organic compounds by developing classification-based RASAR (c-RASAR) models. A diverse, previously curated data set was collected from the literature for which 2D descriptors were computed. The extracted essential features were then used to develop a classification-based linear discriminant analysis (LDA) QSAR model. Furthermore, from the read-across-based predictions, RASAR descriptors were calculated using the basic settings of the hyperparameters for the Laplacian Kernel-based optimum similarity measure. After feature selection, an LDA c-RASAR model was developed, which superseded the prediction quality of the LDA−QSAR model. Various other combinations of RASAR descriptors were also taken to develop additional c-RASAR models, all showing better prediction quality than the LDA QSAR model while using a lower number of descriptors. Various other machine learning c-RASAR models were also developed for comparison purposes. In this work, we have proposed and analyzed three new similarity metrics: g m _class, s m 1 , and s m 2 . The first one is an indicator variable used to generate a simple univariate c-RASAR model with good prediction ability, while the remaining two are similarity indices used to analyze possible activity cliffs in the training and test sets and are believed to play an important role in the modelability analysis of data sets.

Ensemble Quantitative Read-Across Structure–Activity Relationship Algorithm for Predicting Skin Cytotoxicity

“…Although direct comparison between our study and the previous study is not feasible since they modeled the classification-based skin-sensitizing agents. We can observe that the scheme of combination between RA and QSAR model can increase predictive accuracy by approximately 0.05 or 5% compared to the baseline QSAR model . Similarly, in our study, we developed a novel RA model that functioned by similarity-based weighting of outcomes, where similarity coefficients are optimizable power-transformed.…”

“…A recent study by Banerjee and Roy also employed the combination of RA and QSAR methods for a skin sensitization classification model . They invented novel similarity coefficients derived from the physicochemical descriptors coupled with distance-based similarity between the maximum positive and negative classes for model construction.…”

Application of machine learning‐based read‐across structure‐property relationship (RASPR) as a new tool for predictive modelling: Prediction of power conversion efficiency (PCE) for selected classes of organic dyes in dye‐sensitized solar cells (DSSCs)

“…Therefore, the computation of the RASPR descriptors for the test/query set is based on the similarity to the close congeners (source/training compounds) – a Read‐Across‐derived concept. Therefore, the computation of the RASPR descriptors for the test set can be termed “Prediction‐inspired intelligent training” [65], which probably explains why there is an enhancement in the external predictivity. We have made the best model for each data set available (along with sample data sets and instructions) hosted via the site https://sites.google.com/jadavpuruniversity.in/dtc-lab-software/home/pce-prediction-tool.…”

Read-across-based intelligent learning: development of a global q-RASAR model for the efficient quantitative predictions of skin sensitization potential of diverse organic chemicals

“…17,29 Till now, our group has proposed a total set of 18 different RASAR descriptors, two of which can potentially estimate the modelability of a given dataset. 7,30 Some previous studies 7, 30 , also indicate that RA function, g m, and Avg.Sim are three of the most important RASAR descriptors for regression-based RASAR models (q-RASAR).…”