Christian L. Müller

PI in #statistics #datascience #microbiome research @BioDataSc, Munich/New York, @HelmholtzMunich, @LMU_Muenchen, @FlatironCCM, @SimonsFdn, views are my own.

[Christian L. Müller @microbionaut]

We @BioDataSc have open positions for PhD/Postdocs at the interface of statistics, optimization, and microbiome data analysis #stats, #optimization, #microbiome, #datascience, #oceanmicrobiome #gutmicrobiome @CompHealthMuc @LMU_Muenchen; PM me or send an email for more info!!

[NOAA Ocean Exploration @oceanexplorer]

The ocean is full of secrets, and #eDNA is helping scientists solve them. @oceanexplorer has released the 1st eDNA data from expeditions on #okeanos, yielding a wealth of information about life in the deep sea.

Learn more about the data release: https://t.co/NQtWMKlh6i

[Christian L. Müller @microbionaut]

@lpachter Excellent read, thanks!

[Chris Combs (iterative design enjoyer) @DrChrisCombs]

ARTEMIS II AS SEEN BY THE OFFICIAL NASA CESSNA

THIS IS THE BEST LAUNCH VIDEO AND IT ISN'T CLOSE

[deep Manifold @BetaTomorrow]

The reason neural networks still feel like a black box is largely historical: the field has treated learning as a forward problem: optimization over parameters given an objective.

But learning is fundamentally an inverse problem. The model is not solving a predefined governing equation; it is reconstructing a solution geometry from data constraints. When viewed this way, training becomes the numerical search for stable fixed points under weak boundary conditions induced by data.

From this perspective, the mathematics behind neural networks is not mysterious. It reduces to a fairly classical stack: piecewise manifolds (geometry of representation), fixed-point theory (iteration and convergence), and calculus (iterated integrals along intrinsic pathways).

What appears as “emergent intelligence” is simply the behavior of a high-dimensional numerical system solving an inverse problem under massive data constraints. The system is only mysterious if we insist on interpreting it as a forward optimization process rather than as a geometric inverse computation.

see more on https://t.co/sE8yVgJodn

[bioRxiv Bioinfo @biorxiv_bioinfo]

Target-driven optimization of feature representation and model selection for microbiome sequencing data with ritme https://t.co/T0agEvhnaw #biorxiv_bioinfo

[bioRxiv Bioinfo @biorxiv_bioinfo]

Orchestrating Microbiome Analysis with Bioconductor https://t.co/y5Vrxngn6H #biorxiv_bioinfo

[Ernest Ryu @ErnestRyu]

I used ChatGPT to solve an open problem in convex optimization.

*Part I*

(1/N)

[Sundar Pichai @sundarpichai]

Google Earth AI, our collection of geospatial AI models and datasets, is expanding globally and adding new capabilities. That includes Geospatial Reasoning, powered by Gemini, which automatically connects different Earth AI models - like weather forecasts, population maps + satellite imagery - to answer complex questions. 

We’re also bringing new Earth AI models to Gemini capabilities in Google Earth, which make it easy to instantly find objects and discover patterns from satellite imagery. For example, analysts could spot harmful algae blooms that could impact drinking water supply, and issue warnings.

[nature @Nature]

Nature research paper: Isolation, engineering and ecology of temperate phages from the human gut

https://t.co/PmpwZjgNh7

[nature @Nature]

Researchers are turning to AI to help design the next generation of antibiotics to combat rising antimicrobial resistance

https://t.co/5biDmcGdRG

[Microbiome Virtual International Forum @MicrobiomeVIF]

Season 5 is going to start soon!

Check out the program and
feel free to join the keynotes by presenting your work–abstract submission is open: https://t.co/i6eJcbnbte

[Christian Kost @KostChristian]

The position of a senior postdoc (3 +3 years) in the field of theoretical biology is available in my group.

The post provides the opportunity to closely interact with experimentalists and develop own research projects.

Please RT.

Details 👇:
https://t.co/P9M1AOL0zv

[BrochadoLab @AnaRitaBrochad1]

Finally out! A small screen helps us understanding a little more about how E. coli regulates major transporters across chemical stimuli, and reveals how caffeine induces phenotypical antibiotic resistance. Huge thanks to @andrenmateus and @microbionaut
https://t.co/eCxlxU4dXb

[Phys.org @physorg_com]

#Caffeine alters gene regulation in E. coli, reducing antibiotic uptake and weakening drug effectiveness. Dietary components may influence bacterial resistance mechanisms and impact treatment outcomes. @PLOSBiology https://t.co/Lh9ThZCPtc https://t.co/qJsp1WtDtu

[Johannes Brandstetter @jo_brandstetter]

General relativity 🤝 neural fields

This simulation of a black hole is coming from our neural networks 🚀

We introduce Einstein Fields, a compact NN representation for 4D numerical relativity. EinFields are designed to handle the tensorial properties of GR and its derivatives.

[Christian L. Müller @microbionaut]

This!

[Roberto Olayo Alarcon @Scietwas]

Excited that our work "Pre-trained molecular representations enable antimicrobial discovery" has been published in @NatureComms !

Can we leverage unlabeled chemical structures to find new inhibitors of bacterial growth? The answer is YES

Check it out: https://t.co/qYlticx7VO

[BioDataScience Lab @BioDataSc]

Please, meet @Scietwas (Roberto), @SharmaLab1, @AnaRitaBrochad1 and Victoria - a data generation 🤖 from the clip.
They expect to have data for over 200k experiments for
#StressRegNet project which is a part of @bayresq ’s effort to tackle #antibiotic resistance💊💪.

[Roberto Olayo Alarcon @Scietwas]

This work was an awesome interdisciplinary collaboration between @BioDataSc and the @AnaRitaBrochad1 and @SharmaLab1 groups! Many thanks to @bayresq for their support! @LMU_Muenchen @HelmholtzMunich

[Roberto Olayo Alarcon @Scietwas]

Excited that our work "Pre-trained molecular representations enable antimicrobial discovery" has been published in @NatureComms !

Can we leverage unlabeled chemical structures to find new inhibitors of bacterial growth? The answer is YES

Check it out: https://t.co/qYlticx7VO

[Christian L. Müller @microbionaut]

Thanks for the shoutout; this was a great collaborative effort of @BioDataSc , @SharmaLab1 Brochadolab with Medina Feldl and @Scietwas taking the lead on this…

[Cameron Thrash @jcamthrash]

Statistical end-to-end analysis of large-scale microbial growth data with DGrowthR https://t.co/RYFaDuEsZT