FitMultiCell @FitMultiCell

❓How can we speed up ABC on parallel infrastructure? 📜Happy to present our preprint on look-ahead sampling - a wall-time minimizing parallelization strategy for approximate Bayesian computation. 🧵 --> arxiv.org/abs/2305.00506

Multiscale models are helpful to understand the processes determining tissue dynamics. However, inferring such models is difficult. Here is our new #preprint with @yannik_schaelte, @morpheus_lab @JanHasenauer and many other great collaborators. Manuscript: biorxiv.org/content/10.110…

How we can massively parallelize likelihood-free parameter inference for biological multi-scale systems? Happy to share our NIC proceedings at: juser.fz-juelich.de/record/917144 with the help of @yannik_schaelte ,@JanHasenauer and many others from @FitMultiCell consortium.

Enabling mechanistic understanding of infectious diseases using invertible neural networks - Lots of great progress + discussions at our #EMUNE meeting in Heidelberg. 🪅 w/ Ullrich Köthe, Lutz Brusch, @JanHasenauer, Frederik Graw, Noemi Castelletti, ... emune-dev.github.io

📜How can we handle missing data, which are ubiquitous in experimental studies, when performing amortized parameter inference via neural posterior estimation? Happy to present our #preprint with Zijian Wang and @JanHasenauer. ➡️Manuscript: biorxiv.org/content/10.110…

Very much enjoyed this 🌞summer school at @HCM_Bonn! 👩‍🏫Amazing lectures on stochastic models and inference; 👩‍🎓awesome contributed presentations; 🔧practical tool sessions; 🎇great discussions sparking up insights and coops; 👥inspiring panels; 🏃‍♀️fun hikes. hsm.uni-bonn.de/events/hsm-sch…

keyboard_arrow_left Previous keyboard_arrow_right Next

Yannik Schälte @yannik_schaelte

3 years ago

Very much enjoyed this 🌞summer school at @HCM_Bonn! 👩‍🏫Amazing lectures on stochastic models and inference; 👩‍🎓awesome contributed presentations; 🔧practical tool sessions; 🎇great discussions sparking up insights and coops; 👥inspiring panels; 🏃‍♀️fun hikes. hsm.uni-bonn.de/events/hsm-sch… https://t.co/Bv89B0bpet

1 7 17 0 0

Download Image

📜 Here is our paper on #pyABC, published in @JOSS_TheOJ. ➡️Paper: doi.org/10.21105/joss.… ➡️Tool: pyabc.readthedocs.io/en/develop 🧵 We describe various updates to our likelihood-free parameter inference tool.

JOSS @JOSS_TheOJ

3 years ago

📜 Here is our paper on #pyABC, published in @JOSS_TheOJ. ➡️Paper: doi.org/10.21105/joss.… ➡️Tool: pyabc.readthedocs.io/en/develop 🧵 We describe various updates to our likelihood-free parameter inference tool.

0 2 8 0 0

🎒Summer school: 𝐈𝐧𝐯𝐞𝐫𝐬𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦𝐬 𝐟𝐨𝐫 𝐦𝐮𝐥𝐭𝐢-𝐬𝐜𝐚𝐥𝐞 𝐦𝐨𝐝𝐞𝐥𝐬 💬With: Linda Petzold, Christiane Fuchs, @dennisprangle, @StefanEngblom, @A_Hellander ⏰August 22-26 📌@HCM_Bonn ✏️Details+register: hcm.uni-bonn.de/events/eventpa…

🧑‍💻 psst here is our new #pyABC #preprint quickly wrapping up various updates & new features! 🥧🔤 arxiv.org/abs/2203.13043 Thanks esp. to Emmanuel Klinger, @EmadAlamoudi , @JanHasenauer, and many users and collab partners!

📜 Excited to share our new preprint on accounting for informativeness in heterogeneous data via inverse #ML models in likelihood-free #ABC parameter inference! ⬇️🧵 1/n biorxiv.org/content/10.110… @JanHasenauer @CompHealthMuc @LIMES_Bonn @FitMultiCell

Happy to release #pyABC 0.12! 👶 github.com/icb-dcm/pyabc 🧵Featuring a bunch of new features:

Here is our latest preprint on robust adaptive distances for ABC on outlier-corrupted data, based on a method by @dennisprangle. All methods are available in #pyABC. @EmadAlamoudi @JanHasenauer. ➡️ Preprint: biorxiv.org/content/10.110… ➡️ Example notebook: pyabc.rtfd.io/en/develop/exa…

#pyABC 0.11.0 is out! github.com/ICB-DCM/pyABC 🎉 Featuring (esp., see changelog): - robust adaptive distances - speed-ups (><50%) by improved parallelization and more efficient transitions - overhauled logging - better #PEtab support - many fixes and improvements

Happy to share our preprint led by @sisyga91. Bayesian inference and a network-based epidemiological model combining local and random transmissions suggest reducing out-of-cluster contacts to reduce #COVID19 spread, inducing a linear instead of exponential regime (R≈1).

Simon Syga @sisyga91

5 years ago

Happy to share our preprint led by @sisyga91. Bayesian inference and a network-based epidemiological model combining local and random transmissions suggest reducing out-of-cluster contacts to reduce #COVID19 spread, inducing a linear instead of exponential regime (R≈1).

0 0 8 0 0

Looking forward to present our work on simulating and parameterizating computational models of multi-cellular processes (@FitMultiCell ) today at #IbSB2020 at 14:40 CET 🙂.

Check out the @EBItraining virtual course on 'Mathematics of life' with spatial modelling and reproducible modelling. Including a Morpheus tutorial!

EMBL-EBI Training @EBItraining

5 years ago

Check out the @EBItraining virtual course on 'Mathematics of life' with spatial modelling and reproducible modelling. Including a Morpheus tutorial!

0 16 45 0 5

Some compiled information on the recent project on exact ABC inference with noise!

Yannik Schälte @yannik_schaelte

5 years ago

Some compiled information on the recent project on exact ABC inference with noise!

1 2 17 0 1

Download Image

Excited to see my first paper on efficient ABC parameter inference with noisy data out in @OUPBioinfo #ismb2020! #bioinformatics @JanHasenauer @FitMultiCell academic.oup.com/bioinformatics…

4/4 In our latest (recently accepted) work, we show how ABC can actually give efficient exact inference for models with noisy measurements. Here is a little "graphical abstract". ➡️ For more info, see the preprint biorxiv.org/content/10.110…. #YoungScientistsHMGU

3/4 We @JanHasenauer lab @ICBMunich developed the #HPC ready tool #pyABC (github.com/icb-dcm/pyabc), with e.g. self-tuned distances and proposal kernels. Together with @morpheus_lab we develop the @FitMultiCell platform tailored to multi-cellular systems. #YoungScientistsHMGU