The magnificent Robert interviewed me for his Artificial Intelligence podcast -- online tomorrow (May 11) here: aipod.de

AIDD project organized its fifth School at AstraZeneca during the first week of July. The top scientists of the company and invited speaker provided excellent overview of the progress of AI to speed up drug discovery. Selected lectures will be posted at ai-dd.eu/lectures

#ISMBECCB2023 is on! Exciting talk on ancient DNA by Janet Kelso.

Why additive models seem optimal in genetic risk modeling, while we expect nonlinear relationship? If you are interested check our poster at B-234 #ISMBECCB2023

We present our work on optimal transport for domain adaptation in liquid biopsies today on #ISMBECCB2023, Poster B-310. If you are interested, join us!

Our work on identifying clusters of dynamical system behaviour is presented today at #ismbeccb2023, Poster C-412.

The AiChemist project has officially started! This boundary-pushing #phd programme brings together experts in #AI, #chemoinformatics and #chemistry from #academia and #pharma to train future #XAI specialists in the life sciences. aichemist.eu/openpositions --> apply until 10.09!

It was a pleasure to welcome Ana Sanchez-Fernandez in Leuven, where she gave a talk about her work on Representation learning based on cell painting microscopy images. #AIDDOne

It was a pleasure to welcome Paula Torren Peraire in Leuven, where she gave a talk about her work on the effect of single-step retrosynthesis models on multi-step planning performance. https://bsky.app/profile/aidd.bsky.social

You can also find us on sigmoid.social as biolearning.

Our paper (accepted by #CVPR2024) is now available as a preprint on arXiv.org: (arxiv.org/abs/2404.01743). We introduce the first model to perform OCSR with atom-level entity detection with only SMILES supervision. Code is available on GitHub (github.com/molden/atomlenz).

Together with ENNS, @aiddone and https://bsky.app/profile/aichemist.bsky.social, we announce the Tox24 Challenge ochem.eu/static/challen… to predict the in vitro toxicity of compounds tested for activity against transthyretin by U.S. EPA Read more: pubs.acs.org/doi/10.1021/ac… Submit your model by AUGUST 31!

It was a pleasure to have Igor Tetko from Helmholtz Munich presenting us a detailed review of challenges and Advanced ML solutions in Drug Discovery https://bsky.app/profile/aidd.bsky.social https://bsky.app/profile/aichemist.bsky.social

Meet us on #2024, Martijn Adam Arany || mstdn.science/@adamgld

Please join us today afternoon on #CVPR24 poster section, number 297. To hear about our CV method for extracting chemical knowledge from papers and patents.

Exciting keynote by Prof. David Baker on de-novo protein design #CVPR2024

If you are curious how different #uncertainty estimation methods perform when they are tested on NN models trained on large scale industrial #pharma data, visit our poster in room A8 (AI4Science) from 16:15 #ICML2024 @RosaFriesi Adam Arany || mstdn.science/@adamgld Emma Svensson https://bsky.app/profile/aidd.bsky.social

Second talk is by Rosa Friesacher @RosaFriesi from AstraZeneca and KU Leuven speaks about calibration of uncertainty for activity predictions in classification models using time-split validation experiments.

Daniele Raimondi from our lab gave a highlight talk on the ballance of dataset size and model complexity in genome interpretation on #ECCB2024. You can read the original work at genomebiology.biomedcentral.com/articles/10.11…

Andras Formanek, from our lab, will present tomorrow 20th at #ICANN2024 (14:50, Aula Magna) on Model Based Clustering of Time Series Utilizing Expert ODEs. Paper: link.springer.com/chapter/10.100…