You are here:
Extending the Generalised Read-Across approach (GenRA): A systematic analysis of the impact of physicochemical property information on read-across performance
Citation:
Helman, G., I. Shah, AND G. Patlewicz. Extending the Generalised Read-Across approach (GenRA): A systematic analysis of the impact of physicochemical property information on read-across performance. Computational Toxicology. Elsevier B.V., Amsterdam, Netherlands, 8:34-50, (2018). https://doi.org/10.1016/j.comtox.2018.07.001
Impact/Purpose:
This is a systematic evaluation to investigate the impact of physicochemical similarity on read-across performance. In traditional read-across, physicochemical similarity is taken into account on a case by case basis - but the extent to which it impacts an assessment has not been quantified on a systematic and generalization manner. As part of our ongoing evaluations on GenRA, we have considered physchemical similarity as a context to refine and extend the baseline structural similarity approach.
Description:
Read-across is a popular data gap filling technique used within category and analogue approaches in regulatory hazard and risk assessment. Recently we developed an algorithmic, automated approach called Generalised Read-Across (GenRA) (Shah et al., 2016) which makes read-across predictions of toxicity effects using a similarity weighted average of source analogues characterised by their chemical and/or bioactivity descriptors. The GenRA approach served as a first step in establishing a baseline for read-across performance. The default GenRA approach relies on identifying source analogues relative to a target substance that are structurally similar based on chemical fingerprints and computing an activity score to estimate presence or absence of in vivo toxicity effects. In previous GenRA applications, identification and evaluation of analogues did not include information on toxicity effects of interest nor other considerations such as similarity in metabolism, reactivity or bioavailability. This study investigated the role that physicochemical property information plays in altering the local neighbourhood of source analogues as well as read-across predictive performance. Two approaches were evaluated: 1) a filtering approach which restricted structurally related analogues based on their physicochemical properties, and 2) a search expansion approach which included additional analogues on the basis of combined structural and physicochemical similarity. Filtering source analogues was sensitive to the choice filter threshold and could adversely impact the read-across predictions derived. On the other hand, search expansion performed at least as well as a purely structural read-across and markedly improved read-across performance by 5% on average for 10 organs (20%) of organs. We summarise the overall impact physicochemical information plays on GenRA performance and highlight the impacts observed when adjusting similarity threshold using a case study substance. These analyses demonstrate the possibility of codifying physicochemical property information, an important consideration to inform bioavailability, in a systematic manner to make reproducible and objective read-across predictions.
URLs/Downloads:
DOI: Extending the Generalised Read-Across approach (GenRA): A systematic analysis of the impact of physicochemical property information on read-across performance