Antoine Szatkownik
Interested in patterns
Machine Learning – Computational Biology
Paris
2022-2025
PhD, Designing neural networks tailored to population genetics, Laboratoire Interdisciplinaire des Sciences du Numérique (LISN), Université Paris-Saclay, ED STIC, Supervisors: Flora Jay and Guillaume Charpiat.
2020-2022
Master, Specialization in Bioinformatics and Modeling (BIM),
Sorbonne Université.
2018-2020
Bachelor, Fundamentals and Applied Mathematics (MFA),
Université Paris-Sud.
2016-2018
Bachelor, Bidisciplinary in Mathematics and Physics,
Université de Montréal.
2012-2015
Lycée Janson-de-Sailly.
2022
(6 months)
Internship, Master 2 thesis, at Robert Koch Institute, Berlin, Supervisors: Hugues Richard, Elodie Laine. Development of a computational method exploiting the interplay between gene duplication and alternative splicing to learn about the determinants of molecular recognition specificity.
2021
(4 months)
Master 1 thesis at The Laboratory of Computational and Quantitative Biology (LCQB), Supervisors: Elodie Laine, Hugues Richard. Discovery of protein specificity signatures in evolutionary splicing graphs.
2023-2024
• Introduction to machine learning - 3rnd year BCompSc - Lecturer: François Landes - 24h - Teaching assistant
• Deep Learning in Practice - M2 at MVA ENS + CentraleSupelec + DSBA ESSEC - Lecturer: Guillaume Charpiat - 24h - Teaching assistant
2022-2023
Introduction to machine learning - 3rnd year BCompSc - Lecturer: François Landes - 24h - Teaching assistant
Szatkownik, A., Furtlehner, C., Charpiat, G., Yelmen, B. Jay, F.. . Towards creating longer genetic sequences with GANs: Generation in principal component space. Proceedings of the 18th Machine Learning in Computational Biology meeting, in Proceedings of Machine Learning Research 240:110-122 2024
Yelmen, B., Decelle, A., Boulos, LL., Szatkownik, A., Furtlehner, C., Charpiat, G., Jay, F. Deep convolutional and conditional neural networks for large-scale genomic data generation, PLoS Comput Biol 19(10): e1011584,2023, doi: 10.1371/journal.pcbi.1011584
Szatkownik, A., Javier Zea, D., Richard, H., Laine, E. Building alternative splicing and evolution-aware sequence-structure maps for protein repeats. Journal of Structural Biology, 2023, doi: 10.1016/j.jsb.2023.107997
2025
Latent generative modeling of long genetic sequences and their usefulness for local ancestry inference, Séminaire MaIAGE, INRAE Jouy-en-Josas, talk
2024
Diffusion-based artificial genomes and their usefulness for local ancestry inference & HyenaLM for LAI, HFSP, talk
Assessing usefulness of artificial genomes via local ancestry inference, LEGEND: Machine Learning for Evolutionary Genomics Data, Heraklion, talk
Towards creating longer genetic sequences with GANs: Generation in principal component space, NLDL, Tromsø, poster
2023
Towards creating longer genetic sequences with GANs: Generation in principal component space, MLCB, Seattle, poster
Towards creating longer genetic sequences with GANs: Generation in principal component space, LEGO, Lille, talk
Towards creating longer genetic sequences with GANs: Generation in principal components space, Alphy & AIEM, Grenobles, poster
2022
Alternative splicing modulates the number and composition of similar exonic regions, yr2biomodel : second edition of the Young Researchers' Meeting at Sorbonne Université, Paris, talk
Programming (python, R, bash) - HPC (SLURM), Git - Machine Learning / Deep Learning - Processing, statistical analysis and interpretation of biological data - Writing scientific articles (latex)
Languages: French - English