Sylvain V. Costes

The paper presents a framework for imputing sparse, multimodal space biology datasets and shows that MICE-Elastic and a hybrid strategy preserve biological signal better than other methods.

This paper benchmarks cross-species adaptation of a human retinal model to rodent OCT for age estimation, finding that CNN baselines outperform transformer models in small-sample settings.

The paper uses machine learning to analyze spaceflight effects on mammary tissue, finding age-related differences in molecular pathways linked to breast cancer.

This paper presents a federated learning framework deployed on the International Space Station to train models using data from Earth and the ISS without transferring raw data.

This paper pre-trains a machine learning model on Perturb-Seq data to predict perturbation effects in new datasets, including spaceflight gene expression profiles.

The paper identifies Systemic Genome Correlation Loss as a key biosignature of spaceflight by analyzing transcriptomic data from multiple datasets.

The paper identifies genomic associations with ex vivo DNA damage responses to ionizing radiation in mice, highlighting pathway differences across radiation types and timing.

The paper identifies miR-2392 as a miRNA involved in SARS-CoV-2 infection, showing its role in inflammation, glycolysis, and hypoxia, and develops a miRNA-based therapeutic targeting it.

The paper describes a standardized RNA-Seq pipeline for spaceflight data, enabling reproducible analysis of gene expression changes in astronauts and organisms exposed to spaceflight.

The paper shows that RNAseq analysis of rodent spaceflight experiments is confounded by sample collection and preservation techniques, with preservation effects overshadowing spaceflight effects.