The REACH (Registration Evaluation Authorization and restriction of Chemicals) regulation requires from industries the reporting of hazard data for substances placed on the market. Thanks to the registration procedure initiated in 2007, a large REACH database [1] of well defined (eco)toxicological properties has been created. The REACH-chemical space is defined by more than 21’500 compounds registered in the REACH database. Considering the high number of chemicals and endpoints for which experimental data is available, the ability to visualize this chemical space as well as to profile a compound integrating several key properties at once is a growing need. Here, the data distribution in REACH chemical space was visualized and analysed with the help of 2-dimensional Generative Topographic Map (GTM). Similarly to geographical map, on GTM, each object (compound) is visualized as a datapoint. Moreover, compounds possessing similar properties tend to be located in neighbourhood. The third dimension can be added in order to display a distribution of the given (eco)toxicological property (such-called “property landscape”), which can further be used for property assessment of new compounds projected on the map. We report the universal REACH map which accommodates 11 endpoints, covering environmental fate, and (eco)toxicological properties. This map is able to provide predictions for each property, and demonstrates acceptable predictive performance in cross-validation: balanced accuracy ranges from 0.60 to 0.78. Superposition of different property landscapes allows to delineate the “areas of interest” populated by molecules possessing desirable (eco)toxicological profile.

{"references": ["echem portal: global portal to information on chemical substances (OECD). https://www.echemportal.org/echemportal/", "K. Mansouri, A. Abdelaziz, A. Rybacka, A. Roncaglioni, A. Tropsha, A. Varnek et al., CERAPP: collaborative estrogen receptor activity prediction project, Environ. Health Perspect. 124 (2016), pp. 1023\u20131033.", "K. Mansouri, N. Kleinstreuer, A.M. Abdelaziz, D. Alberga, V.M. Alves, P.L. Andersson et al., Compara: collaborative modeling project for androgen receptor activity, Environ. Health Perspect. (2020).", "Kleinstreuer, Nicole C., et al. \"Predictive models for acute oral systemic toxicity: a workshop to bridge the gap from research to regulation.\" Computational Toxicology 8 (2018): 21-24.", "F. Lunghini, G. Marcou, P. Gantzer, P. Azam, D. Horvath, E. Van Miert and A. Varnek, Modelling of ready biodegradability based on combined public and industrial data sources, SAR QSAR Environ. Res. 31 (2020), pp. 171\u2013186.", "F. Lunghini, G. Marcou, P. Azam, D. Horvath, R. Patoux, E. Van Miert and A. Varnek, Consensus models to predict oral rat acute toxicity and validation on a dataset coming from the industrial context, SAR QSAR Environ. Res. 30 (2019), pp. 879\u2013897."]}