Constantin Ahlmann-Eltze (@const-ae.bsky.social)

Postdoc at UCL with @JL_Reading. Previously at EMBL working with @wolfgangkhuber. Biostats, R, cancer immunology
Find me on 🦋 at

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

@NadigAjay Thanks :) I discussed the idea a bit in https://t.co/RM38mZeZ3b

However, such a test would require some additional statistical groundwork. I would love it if someone picked this up :)

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

After 4y in the making, I am super excited that my main PhD project is published 🎉🥳🎉🎉🥳

https://t.co/LCL2Dj1SOK

LEMUR is a tool to analyze multi-condition single-cell data and model differential expression as a continuous function of the cell-state space.

Some highlights⬇️

[Clint Miller @clintomics]

New single-cell analysis framework, LEMUR, disentangles covariates and latent cell states from multi-condition data to predict counterfactual gene expression and identify cell neighborhoods with similar DEGs without clustering. https://t.co/Dktxxek5C7 @NatureGenet @wolfgangkhuber

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

For the Monday morning crowds :)

Also, take a look at my blog post, in which I compress the full lemur package into 100 lines of code and explain the core algorithm.

https://t.co/c6ajj3DMn7

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

After 4y in the making, I am super excited that my main PhD project is published 🎉🥳🎉🎉🥳

https://t.co/LCL2Dj1SOK

LEMUR is a tool to analyze multi-condition single-cell data and model differential expression as a continuous function of the cell-state space.

Some highlights⬇️

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

Big shout out to everyone involved: my supervisor @wolfgangkhuber and everyone else who helped along the journey: @OliverStegle @s_anders_m @zauggj @LSaund11 https://t.co/EQCtrE1n9Q

And lastly, this link should work even without a Nature subscription: https://t.co/2cyIXnFJQv

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

Finally, we demonstrate LEMURs usefulness on

- a treatment vs. ctrl dataset from glioblastoma,
- Zebrafish developmental time course, and
- a spatially resolved Alzheimer dataset

where we find intriguing DE patterns

[Aino-Maija Leppä @amleppa]

Extremely proud to see my PhD project exploring intra-tumor heterogeneity in complex karyotype AML finally out in @NatureGenet ! 🎉👩🏻‍🔬

They say that 10,000 hours make you an expert, so I guess after more than five years of work I’m half-way there 😅 ⬇️1/5
https://t.co/xe0WrtSRse

[Hadley Wickham @hadleywickham]

The #rstats community on bluesky seems to be really picking up steam, so you can also find me over there: https://t.co/NuxyZJ92UF

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

@kasparmartens Very nice work. I like your mean-perturbation-effect baseline, and the LLM embeddings from the sequence and text are a cool idea. I will update the Discussion to incorporate this :)

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

There's a lot of excitement about foundation models and their ability to learn biology 🧬💻

But current tools for perturbation prediction perform worse than simple linear models! We need more careful benchmarking to make progress.

https://t.co/lTJM7ghk2r

[Charlotte Bunne @_bunnech]

🚀Interested in doing a PostDoc in my lab?
Consider applying for a🇨🇭SNSF Swiss Postdoctoral Fellowship 2024. If you have any questions, join the information event tomorrow via Zoom (register via https://t.co/QsTKkraRIr) or reach out to us via https://t.co/KXW2tpttqy!

🧬Who should apply? Check that you fulfill the requirements. And: We are in particular interested in candidates with background in computational biology or machine learning applied to the biomedical domain!

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

@science_wallet That sounds exciting, can't wait to see what you are cooking up :)

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

Great thread by Anshul, one of the pioneers of deep learning for biology, summarizing our preprint!

[Anshul Kundaje @anshulkundaje]

https://t.co/QCtOIaybOj

Nice benchmark of single cell "foundation models" (scGPT, scFoundation) and GEARS (a GNN model) further hyped as "virtual cell models" against linear baselines on perturbation prediction.

Long-story short: they can't beat the linear baselines. 1/

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

@pan_cancer @Alejandro__TL replied on Github and clarified that Prophet in its current form is not intended for predicting perturb-seq experiments, but that they are considering adding that feature in the future.

[Constantin Ahlmann-Eltze (@const-ae.bsky.social) @const_ae]

@pan_cancer Yes, it would be great to include Prophet in the benchmark. I just asked how to best apply the software to genetic perturbation data (https://t.co/ShvI0woELK) because I don't quite understand yet how to use the software :)