Posts on X

Initially we wanted to, but ended up recording and will upload to YouTube tomorrow or so. Worthwhile to check out some of the talks for sure.

@eli_mereu

@fabian_theis @TomAsh_1 @singlecellomics Hi @fabian_theis! Is there a live streaming for talks??

Quoted post

Thanks for a great talk, Peter, as well as big shout of gratitude to all speakers. Really enjoyed the meeting!

@singlecellomics

Captivated audience at #SCOGmunich2019 while @KharchenkoLab introduces the Conos approach and @humancellatlas work on bone marrow and more

Quoted post

Thanks, Wolfgang! Quite a challenge ahead of us, looking forward to working with Munich and Helmholtz colleagues on getting this on track soon. x.com/wolfgangkhuber…

I really like our cell-like digitized scampi. Go #scanpy scampi!

@falexwolf

Since today, Scanpy has a logo with a cellular decomposition of a Scampi in it. 😀 scanpy.rtfd.io

Quoted post

Great poster session in beautiful location at #SCOGmunich2019

True - some definite bias in the crowd given genomics in the title :)

@SavasTay_UChi

@fabian_theis my additions: protein, protein complexes, protein modifications!

Quoted post

Thanks for coming, chairing and talking!

@KharchenkoLab

@fabian_theis kicks off #SCOGmunich2019 in a fancy venue!

Quoted post

Voting on current challenges in single cell analysis at #SCOGmunich2019.

Looking forward to today’s workshop on computational challenges in @singlecellomics here in Munich, with a great lineup of speakers and a Hackathon followup. Program: bit.ly/2WnmUTC

Interesting read and thorough comparison! Title may be a bit off since deep learning is only minor part in this comparison, no?

@UnlearnAI

Do machine learning methods + transcriptomic data = more powerful tools to predict health outcomes? Our collaboration with Pfizer gives a surprising answer: bit.ly/2F8mSJG

Quoted post

Happy Pi day everyone

@MathigonOrg

Super excited for #PiDay tomorrow! Stay tuned for some exciting updates…

Quoted post

More time to submit a paper to this year‘s ICANN at Munich.

@EuNeuralNetSoc

Deadline for paper submission at #ICANN2019ENNS extended to 1st of April 2019. e-nns.org/icann2019/subm… #conference #AI #MachineLearning

Quoted post

Agreed - September is just always packed. But there‘s at least a very fast train connection between Würzburg and Munich..

@kmborgwardt

@fabian_theis @EuNeuralNetSoc Hui, fully overlapping with ECMLPKDD in Würzburg! ecmlpkdd2019.org Great for Bavaria that it hosts both, bad that it‘s the same week...

Quoted post

Bavaria, the most southern part of Germany, is conservative. But we have a fun tradition, 'Politiker derblecken' = making fun of politicians during Lent. This year done by Maxi Schafroth, ending with suggestion that politicians should govern as if they were the governed. #nohate

I agree - indeed in academia the 'solution delivery' part is often underdeveloped, and bringing our results to clinics pretty hard. Guess @HelmholtzMucEn we are also getting better with this but could learn/coop more with companies along these lines.

@ewanbirney

@OliverStegle @fabian_theis I'd also say that "development" and "delivery" is not a company only world (take @emblebi for example - we do *lots* of this, but also large scale cohorts, process and technique optimisation in protocols, engineering for new instruments, eg beamlines, cryo-EM etc)

Quoted post

Yeah, I can see this. Not so sure about the related field of machine learning and AI though, where in parts research in industry is pretty impressive ..

@lpachter

Hypothesis: the fields of genomics and bioinformatics will be just fine despite the recent departure of a handful of guys from academia. x.com/erlichya/statu…

Quoted post

Congrats for getting this ambitious grant - and beautiful logo

@lucaspelkmans

New research initiative in Switzerland. Excited to be part of their ambitious goals: Connecting scales in single-cell biology.

Quoted post

Just wanted to weigh in how right you are!

@jenheemstra

Hard reality: how much of our busyness is fueled by our pride -the group email where we *have* to weigh in -the meeting where our input is *necessary* to reach the right decision -the task that *only* we can do Tough but valuable to admit some things will be just fine without us

Quoted post

Cool visual summary of the paper - please also checkout theislab.github.io/LungAgingAtlas

@lkmklsmn

Our lung aging atlas is out @NatureComms today! Big thx to the reviewers who substantially improved the manuscript. We added a total of 21(!) panels to the main figs (def worth a re-read ;-). Nice collab bw @ICBmunich & @CPCSchool! Check out our interactive webtool!

Quoted post

Great effort and resource, really enjoyed the collaboration!

@SchillerLab

#Lung, #Ageing, #SingleCell Our multi-modal and multi-omics lung aging atlas out today in @NatureComms. @SchillerLab @fabian_theis @lkmklsmn @Eickelberg_MD @labs_mann @HelmholtzMucEn @singlecellomics @humancellatlas bit.ly/2IEpO4c bit.ly/2Eml6mi [1/6]

Quoted post

Surprise attack by Viserion from #GoT at the Singapore Zoo - hope that Rahul @satijalab makes it for his talk @CellSymposia tomorrow.

You are very kind, thank you! Really enjoyed the night zoo venue, too, btw!

@asharmaiisc

Best part of conference is opportunity to meet & interact with your hero’s. I had the opportunity to meet @fabian_theis in person a great scientist & very humble individual.. loved our interactions on trajectories & data integration #ScanPyFan #CSSingleCell

Quoted post

@Prof_Lundberg on Citizen Science for microscopy image classification at Cell meeting at Singapore. Gamers from #eveonline spent 70 working years on this mini game - impressive!

'Deep Work' instead of emails and admin - yes, please!

@MarkGerstein

"Is Email Making Professors Stupid?" is the Q posed by Chronicle.com/interactives/i… . My A: YES. The article has a nice description of the problem with 24/7 connectivity: how the urgent but unimportant crowds out the important but non-urgent

Quoted post

This is Bavaria, governed by a still far too male political party; no awareness. Loved your term 'manel' though, awesome!

@EleZeggini

Interesting discussion on synergies of the UK industrial strategy sector deal on AI and Bavaria’s Digital Programme at the impressive IBM Watson Tower in Munich (shame about the almost-manel though)

Quoted post

‚It now appears that the answers already surround us. It’s the question we don’t know.‘ - this is soo Hitchhiker‘s guide to the galaxy! #42

@N_Rajewsky

i think this is an inspiring article: A Different Kind of Theory of Everything newyorker.com/science/elemen…

Quoted post

The feeling when a tool you have been using for ages has a feature you always needed but never found. (In my case, I just learnt that Keynote can subtract shapes, lovely!)

I guess SZ used the term as heading because of the double use not only for model but also for a measurement device @LickertHeiko‘s lab we showed :)

@dermitzakis

@wibior @fabian_theis @SZ @vicigo @JeanetteErdmann Off the top of my head: "multifactorial diagnosis", "computational diagnosis", "computer-assisted diagnosis"

Quoted post

Agreed! In fact we talked about explainable AI and how to go exactly beyond a predictor, ok eg in segmentation but maybe not for patient treatment. Text in German though, sry..

@dermitzakis

@fabian_theis @SZ @vicigo Great photo! But without wanting to kill your thunder the term “blackbox diagnosis” is both wrong and hugely counterproductive to convince people about these methods.

Quoted post

Today even my son read the newspaper - daddy proudly featured in an @SZ article on #AI in medicine, and 'blackbox-diagnoses'. Author @vicigo totally helped explain to my kids what I do. sz.de/1.4321839