Jesper Madsen's Avatar

Jesper Madsen

@jespergrud.bsky.social

Associate Professor @ University of Southern Denmark. Computational Biologist working with ML to understand transcriptional and genetic regulation of metabolic disease.

64 Followers  |  223 Following  |  7 Posts  |  Joined: 15.11.2024  |  1.6233

Latest posts by jespergrud.bsky.social on Bluesky

Shoutout to funders without whom this work would not have been possible @novo-nordisk.bsky.social @dg.dk

07.10.2025 12:39 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
Disentangling covariate effects on single cell-resolved epigenomes with DeepDive Understanding the effects of individual biological factors from single cell-resolved epigenomic data is hindered by multicollinearity, particularly in human cohorts. We introduce DeepDive, a novel dee...

You can find many more details in the preprint biorxiv.org/content/10.1.... You can also try out DeepDive on your own data. You can find a tutorial on how to get started here: deepdive-tutorial.readthedocs.io/en/latest/. Reach out if you have questions or something does not work. 5/6

07.10.2025 12:39 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Post image

Finally, we apply DeepDive to islets (Sander lab), and dive into beta-cell subtypes linked to T2D. We id TFs differentially active between subtypes and after virtually perturbing disease status. TFs with counterfactual support more often have genetic and functional support! 4/6

07.10.2025 12:39 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Post image

Next, we turn counter-factual prediction and attempt to use the disentangled representations to predict the epigenome in an unseen covariate combination by virtually perturbing the observed covariates in real data. DeepDive outperforms baselines and achieves high accuracy. 3/6

07.10.2025 12:39 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Post image

To evaluate disentanglement, we analyzed collinear covariates in both simulated and real datasets. We found that regular statistical methods fail, as signal from one covariate leaks into another, but DeepDive accurately resolves which signal belongs to which covariate. 2/6

07.10.2025 12:39 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Post image

Excited to share DeepDive, a generative model for disentangling covariate effects in snATAC-seq developed by the amazing @andreasfm.bsky.social. DeepDive enables deeper understanding of a dataset by separating the effects of known and unknown covariates. Here are some highlights. 1/6

07.10.2025 12:39 โ€” ๐Ÿ‘ 4    ๐Ÿ” 3    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Preview
Hepatic stellate cells regulate liver fatty acid utilization via plasmalemma vesicle-associated protein Hansen etย al. show that the hepatic stellate cell-expressed plasmalemma vesicle-associated protein is required for the normal switch from carbohydrate to lipid utilization during fasting. This reveals...

Check out our latest paper now online at Cell Metabolism: Hepatic stellate cells regulate liver fatty acid utilization via plasmalemma vesicle-associated protein. Amazing work by Daniel, Jasmin, Christian, Peter, Oliver from @ravnskjaer_lab, and many more ๐Ÿ†๐Ÿ’โ™ฅ๏ธ
www.cell.com/cell-metabol...

03.03.2025 15:57 โ€” ๐Ÿ‘ 13    ๐Ÿ” 7    ๐Ÿ’ฌ 2    ๐Ÿ“Œ 1
Preview
Single cell-resolved transcriptional dynamics of human subcutaneous adipose tissue during lifestyle- and bariatric surgery-induced weight loss During sustained weight gain, human white adipose tissue undergoes dramatic remodeling that may compromise adipose tissue function and lead to obesity comorbidities such as cardiometabolic disease. Th...

Preprint alert! We used snRNA-seq on adipose tissue from a longitudinal cohort of people undergoing surgury-induced weight loss. True ATLAS team effort

We found dramatic changes in immune and progenitor compartments. Give it a read and let us know what you think.

www.biorxiv.org/content/10.1...

01.02.2025 08:44 โ€” ๐Ÿ‘ 2    ๐Ÿ” 1    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 1

@jespergrud is following 20 prominent accounts