You are here:
Towards systematic read-across using Generalised Read-Across (QSAR 2023)
Citation:
Patlewicz, G. AND I. Shah. Towards systematic read-across using Generalised Read-Across (QSAR 2023). QSAR, Copenhagen, DENMARK, June 05 - 09, 2023. https://doi.org/10.23645/epacomptox.22659157
Impact/Purpose:
N/A
Description:
Generalised Read-Across (GenRA) is a data-driven approach to estimate biological and eco-toxicological properties of chemicals from similar analogues. GenRA attempts to mimic the expert driven read-across workflow for filling data gaps using an interpretable and automated approach based on nearest-neighbours. The read-across prediction is a similarity weighted activity of nearest neighbours based on chemistry and/or bioactivity fingerprints. GenRA is a web-based application that is linked to the EPA CompTox Chemicals Dashboard. GenRA 3.1 was last released in September 2022 with several significant new features including: 1) ability to download the radial plot view and top 100 most similar analogues; 2) the ability to explore physicochemical similarity by exploring the distribution of specific properties across analogues; 3) a network tool to enable an exploration of neighbourhoods of source analogues and compare them across different fingerprint types; 4) prediction of in vivo toxicity potencies rather than binary toxicity predictions from dose values aggregated in ToxRefDB and making use of the genra-py library [1]; and 5) prediction of ToxCast hitcall outcomes – on a per assay basis. The latest version of GenRA, version 3.2 will add the ability to 1) download fingerprints for the top 100 similar analogues, 2) identify analogues on the basis of Analog Identification Methodology (AIM) fingerprints [2] and 3) sort on different options to help prioritise the predictions generated. The GenRA application is available from comptox.epa.gov/genra/. This abstract does not reflect EPA policy.
URLs/Downloads:
DOI: Towards systematic read-across using Generalised Read-Across (QSAR 2023)
QSAR2023_GENRA_POSTER_PATLEWICZ_120423.PDF (PDF, NA pp, 1184.752 KB, about PDF)