Carlo De Donno
Senior ML Scientist at Roche
Basel, Basel-City, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolCarlo De Donno is a Senior ML Scientist based in Basel with eight years of experience applying deep learning and generative models to drug discovery and computational biology. Trained at Technical University of Munich (Dr. rer. nat.) with a strong neuroengineering background, he has led interpretable VAE and perturbation-prediction projects spanning single-cell and bulk data, and has embedded as a data scientist within Roche pRED’s RNAHub designing gene regulatory elements and RNA editors. His work bridges academia and industry—collaborating with FAANG, pharma partners, AWS, and leading scRNA-seq analyses across multiple projects—bringing research-grade methods into production drug programs. Known for pushing graph ML and causal perturbation approaches, he combines rigorous modeling with pragmatic engineering to accelerate early drug discovery.
8 years of coding experience
8 years of employment as a software developer
Dr. rer. nat. Computational Biology and Machine Learning, Dr. rer. nat. Computational Biology and Machine Learning at Technical University of MunichScientific lyceum diploma Scientifico Maxisperimentale, Scientific lyceum diploma Scientifico Maxisperimentale at Scientific LyceumBachelor of Science - BS Bioengineering and Biomedical Engineering, Bachelor of Science - BS Bioengineering and Biomedical Engineering at Politecnico di Torino
Italian, English, Russian, German, Ukrainian
Github Skills (37)
scales10
single-cell-genomics10
multimodal-deep-learning10
autoencoder10
generative10
transcriptomics10
single-cell-analysis10
bioinformatics10
uncertainty10
scanpy10
combinations10
data-science9
python9
machine-learning9
rna-velocity9
Programming languages (3)
Github contributions (5)
theislab/scPoli_legacy
Jul 2021 - Nov 2022
Contributions:61 commits in 1 year 5 months
facebookresearch/CPA
Nov 2021 - Jul 2022
The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.
Contributions:58 commits, 1 PR, 31 pushes in 8 months
drugsdoseinterpretablesingle-cellautoencoder
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.