The on-line course on using ochem.eu for model development, held as part of the 9th Workshop on Advanced Computer-Aided Drug Design 2026 (kfc.upol.cz/9add), was also focused on analyses of the best methods for developing models that won the #Tox24 and overview of results of the #EUOS2025 challenge

Igor Tetko, Barry Hardy, Ola Spjuth and Stefan Kramer. I am answering question of Ola whether human expertise is still required in the era of AI.

OpenTox Virtual Conference 2025 www.opentox.net/events/virtu... had a great selection of speakers covering different aspects of computational toxicology. It was nice to speak there with a critical analysis of the results of #Tox24 challenge: representation learning and consensus models are the best!

#AutumnOnlineTrainingSchool2025 had excellent speakers www.linkedin.com/feed/update/... covering topics from Screening Platforms to AI for Drug Discovery. >230 participants were attending my lecture with overview of #Tox21 #Tox24, Kaggle Solubility & new EUOS2025 doi.org/10.1016/j.sl... Join it now!

The visit of Dalian University of Technology, School of Environmental Science and Technology was great possibility to meet again Prof. J. Chen, learn about their new studies, summarise #Tox24 & promote #EUOS2025 challenges. Well there was one more, a culinary challenge, during the same date...

Igor Tetko gave an on-line seminar saferworldbydesign.com/webinars/56/ with an overview of #Tox24 challenge (see pre-print with S.A. Eytcheson doi.org/10.26434/che...)
The advantages of consensus models, which contributed most of winning models, were discussed. See also linkedin.com/feed/update/...

Consensus Modeling Strategies for Predicting Transthyretin Binding Affinity from Tox24 Challenge Data Transthyretin (TTR) is a key transporter of the thyroid hormone thyroxine, and chemicals that bind to TTR, displacing the hormone, can disrupt the endocrine system, even at low concentrations. This st...

Congratulations Thalita Cirino, @m-iwan.bsky.social from @aichemist.bsky.social and all their coauthors with publication pubs.acs.org/doi/10.1021/... summarising 9 models contributing top predictions of Transthyretin Binding Affinity in #Tox24 Challenge, which is published by the ChemResTox today!

Enhancing Transthyretin Binding Affinity Prediction with a Consensus Model: Insights from the Tox24 Challenge Transthyretin (TTR) plays a vital role in thyroid hormone transport and homeostasis in both the blood and target tissues. Interactions between exogenous compounds and TTR can disrupt the function of t...

The second article describing group winning model of #Tox24 challenge co-organised with @aidd.bsky.social was just published by @pubs.acs.org pubs.acs.org/doi/10.1021/... Congratulations to Xiaolin Pan @xlpan.bsky.social and his co-authors! Do not miss reading about strategies how to win Challenges!

The #Tox24 Challenge was co-organized by @e-nns.bsky.social, @aidd.bsky.social and ChemResTox @pubs.acs.org. More models will be published soon. Stay tuned!

Consensus Modeling for Predicting Chemical Binding to Transthyretin as the Winning Solution of the Tox24 Challenge The utilization of predictive methodologies for the assessment of toxicological properties represents an alternative approach that facilitates the identification of safe compounds while concurrently reducing the financial costs associated with the process. The objective of the Tox24 Challenge was to assess the progress in computational methods for predicting the activity of chemical binding to transthyretin (TTR). In order to fulfill the requirements of this task, the data set, measured by the Environmental Protection Agency, consisted of 1512 chemical substances of diverse nature. This paper describes the model that won the Tox24 Challenge and the steps taken for its further improvement. The Transformer convolutional neural network (CNN) model achieved the best performance as a standalone solution. Meanwhile, a multitask model built on a graph CNN, trained using 11 additional acute systemic toxicity data sets with increased weighting on the TTR binding activity, showed comparable results on the blind test set. The winning solution was a consensus model consisting of two catBoost models with OEstate and Mold2 descriptor sets, as well as two transformer-based models. The improvement of this solution involved adding a fifth model based on multitask learning using the graph CNN method, which led to a reduction in RMSE on the blind test set to 20.3%. The winning model was developed using the OCHEM web platform and is available online at https://ochem.eu/article/162082.

The model that won the #Tox24 Challenge e-nns.org/icann2024/ch... has just been published by the ACS doi.org/10.1021/acs.....

The consensus model used representation learning doi.org/10.1186/s133... and mixture descriptors. Check out doi.org/10.26434/che... for an overview of the other top models.