Interested in #Bioinformatics and #SpatialAnalysis? Check out this #preprint! Klein et al developed a new framework to map single cells across time and space called #moscot & use the #MERFISH #MERSCOPE #Mouse #Liver & #Brain #datasets: hubs.ly/Q01SK-Xm0 #AcademicTwitter

moscot - multi-omic single-cell optimal transport tools github.com/theislab/moscot It can be used for trajectory inference, mapping cells to their spatial organisation, aligning spatial transcriptomics slides, translating modalities prototyping of new OT models in single-cells

Want to learn cell state dynamics beyond RNA velocity? Led by Philipp Weiler & Marius Lange, and with dana_peer, our new CellRank 2 enable fate mapping leveraging pseudotime, gene scores, time points or metabolic labeling in multi-view single-cell data. biorxiv.org/content/10.110…

Presenting mubind, a new method at Fabian Theis lab to learn scATAC-seq using seq features and graph representations e.g. kNN-graphs for cell-cell communication. With it, we can learn motif features linked with local cell transitions #ISMBECCB2023 [1/3] github.com/theislab/mubin…

CellRank 2 interfaces with moscot and moslin to facilitate analyzing spatiotemporal and lineage-traced data, respectively Dominik Klein Giovanni Palla Zoe Piran. See moscot (bit.ly/moscot-biorxiv), moslin (bit.ly/moslin-biorxiv), and tutorial: bit.ly/cr-time. (11/12)

🍏🍏🍏 Apple ML Research in Paris has multiple open research internship positions! 🍎🍎🍎 We are looking for Ph.D. students interested in generative modeling, optimization or uncertainty quantification, with applications to challenging scientific problems. Details below 👇👇👇

Excited that veloVI is out Nature Methods! Led by Adam Gayoso & Philipp Weiler w awesome Yosef Lab, we combine deep generative modeling with a mechanistic single-cell transcriptional model for better RNA velocity analysis with uncertainty quantification. nature.com/articles/s4159…

Is binarization of scATAC-seq data really necessary? The conclusion from our analysis is that a quantitative treatment is in fact beneficial. Now out in Nature Methods! gagneurlab Fabian Theis nature.com/articles/s4159… Many additions since the preprint 👇(1/n)

Join us for a discussion on uncovering new biological insights through the application of Optimal Transport. We'll start with discrete OT and continue with novel Neural OT algorithms using Flow Matching! Can't wait? Feel free to check out moscot-tools.org!

This weekend, more than 1.5 mio (!) people took to the streets against the far-right AfD in Germany - and people should know about it. As in many countries, right-extremism is rising, elections are close, it seemed hopeless. Yet, this wave erupted over night. It is so hopeful.🔥

Check out our work on unbalanced neural optimal transport with numerous real-world applications in both single-cell genomics (trajectory inference, perturbation modeling) as well as in computer vision! #ICLR2024

A variant of mini batch OT flow matching with unbalanced couplings Also very nice to see this in jax: github.com/ExplainableML/… Using: - ott-jax for transport solvers github.com/ott-jax/ott And Patrick Kidger ‘s - equinox for networks - diffrax for ode solvers

Entropic (Gromov) Wasserstein Flow Matching with GENOT ift.tt/3jFSRH9

1/n Happy to finally share our Nicheformer preprint, this was a huge lab effort - special thanks to Anna Schaar and Fabian Theis for making this possible!

Excited to attend my first conference in person #ICLR2024 Join us at the poster session on Wednesday 16:30-18:30 at poster #238! Dominik Klein Théo Uscidda

Also today (May 8th, 4:30pm at Session 4)! Ask Luca Eyring, Dominik Klein and Théo Uscidda on how to utilize unbalanced optimal transport to handle undesired distribution shifts in unpaired translation tasks! With Giovanni Palla, Niki Kilbertus, Zeynep Akata, and Fabian Theis!

Mixed Models with Multiple Instance Learning (MixMIL) received an Oral & Outstanding Student Paper award at AISTATS Conference last week! 🏆 MixMIL enables accurate & interpretable patient label prediction from single-cell data by adding attention to GLMMs.#singlecell #MachineLearning

I am really happy and excited to see CellRank 2 out in Nature Methods (nature.com/articles/s4159…)! Nature Methods Helmholtz Munich | @HelmholtzMunich Computational Health Center TU München Marius Lange Fabian Theis dana_peer (1/16)

Happy to present muBind’s preprint and latest release (v0.2.1)! This is our deep-learning model for DNA-regulatory inference during cell transitions in single-cell genomics [1/10]. Fabian Theis code: github.com/theislab/mubind preprint: biorxiv.org/content/10.110…

1/8 Happy to share Prophet, done with Yuge Ji and Fabian Theis! We have trained a single model across more than 4.7M public cell experiments to predict phenotypic responses (IC50, Cell Painting morphology features, RNA expression, ...). One single model to predict everything.