@atzkenneth

AI Research Scientist @Roche. PhD AI/ML @ETH. 👨🏻‍💻🧬

 Zurich, Switzerland
kennetyh.github.io
Joined October 2013
  • Tweets321
  • Following1.9K
  • Followers582
21Media
Happy to share that I could defend my PhD thesis during my doctoral examination at @ETH. Heartfelt thanks to @clemensisert for the creation of this beautiful hat 🎓 @ETH_en @ETH_DCHAB #drugdiscovery #machinelearning #compchem #deepLearning
5
39
2,125
Happy to share our latest #preprint. 🧬👨🏻‍💻🎉 We introduce a curated training data set of 13.5k C–H alkylation reactions and show how #reactionprediction can yield a streamlined approach for accelerating #drugdiscovery. 🔎💊 Paper: doi.org/10.26434/chemrxiv-20… @Roche @ETH_BSSE
1
3
113
Expediting hit-to-lead progression in drug discovery through reaction prediction and multi-objective molecular optimization #machinelearning #compchem dx.doi.org/10.26434/chemrxiv…

Expediting hit-to-lead progression in drug discovery through reaction prediction and multi-object...

The rapid and economical synthesis of novel bioactive compounds remains a significant hurdle in drug discovery efforts. This study demonstrates an integrated medicinal chemistry workflow that...

chemrxiv.org
1
9
540
With the scarcity of high-quality, leakage-free benchmarking datasets in the chem-bio space, it's really exciting to see this released today!
Luca Naef@NaefLuca19 Jul 2024
💥 Introducing PINDER & PLINDER With existing evals saturating without clear advances in real life downstream tasks, current progress in AI x Bio is primarily rate-limited by better datasets & evals. In two back-to-back preprints, we address this via fantastic academic-industry collaborations by @vant_ai @NVIDIA & the group at @TorstenSchwede that co-organizes CASP, @MIT: Read the release blog posts at: plinder.sh/blog/plinder & pinder.sh/blog/pinder - 🧵(1/6)
3
21
1,658
Congratulations to Sylke Poehling! I am excited about the potential impact of Sylke as Professor of Practice in the Department @ETH_BSSE in enhancing collaboration, scientific and educational exchange between @ETH Zurich and Roche pRED. Read more: go.roche.com/fr5a3
1
1
13
4,706
xLSTM is out -- putting LSTM networks on steroids to become a more than serious LLM competitor. How? Via exponential gating and enhanced (cell state) memory capacities. Does it work? Oh, yeah 🚀🚀 arxiv.org/abs/2405.04517

xLSTM: Extended Long Short-Term Memory

In the 1990s, the constant error carousel and gating were introduced as the central ideas of the Long Short-Term Memory (LSTM). Since then, LSTMs have stood the test of time and contributed to...

arxiv.org
6
72
7
353
40,597
Kenneth Atz retweeted
ETH Zürich@ETH24 Apr 2024
ETH-Forschende haben ein Computerverfahren entwickelt, mit dem sich pharmazeutische Wirkstoffe effizient aus der 3D-Oberfläche eines Proteins herstellen lassen - das hat das Potenzial, die Forschung zu revolutionieren. 💊 @atzkenneth #Arzneimittel #KI brnw.ch/21wJ7rF
7
6
1,936
Kenneth Atz retweeted
ETH Zurich@ETH_en24 Apr 2024
Chemists at ETH created a computer process to produce pharmaceutical ingredients efficiently from a protein's 3D surface, which has the potential to revolutionise research. 💊 @atzkenneth #DrugDevelopment #AI brnw.ch/21wJ7rC

AI designs new drugs based on protein structures

A new computer process developed by chemists at ETH Zurich makes it possible to generate active pharmaceutical ingredients quickly and easily based on a protein’s three-dimensional surface. The new...

ethz.ch
1
3
14
3,261
Kenneth Atz retweeted
D-CHAB@ETH_DCHAB24 Apr 2024
A new computer process developed by chemists led by Gisbert Schneider & Kenneth Atz @ETH_en makes it possible to generate active pharmaceutical ingredients quickly based on a protein’s three-dimensional surface. Learn more on: doi.org/10.1038/s41467-024-4… ethz.ch/en/news-and-events/e…
1
1
10
987
Prospective de novo drug design with deep interactome learning #machinelearning #compchem nature.com/articles/s41467-0…

Prospective de novo drug design with deep interactome learning

Nature Communications - The use of data-driven generative models for drug design is challenging due to the scarcity of data. Here, the authors introduce a “zero-shot" generative deep...

nature.com
2
18
58
3,437
Great catching up in Paris. 🇫🇷🥐🍷Approx. mid-way between Zurich and Cambridge. @clemensisert @josejimlun Croissants at Montmartre instead of quantum chemistry. #compchem #MachineLearning #sciencetwitter – at Paris, France
12
327
New preprint! In this project spearheaded by Anthony Marchand and @St_Buckley we designed new chemically-induced protein interactions and experimentally validated three binders targeting Bcl2:Venetoclax, DB3:Progesterone and PDF1:Actinonin complexes. biorxiv.org/content/10.1101/…
5
35
6
128
29,121
It was fantastic presenting our work from @ETH_en and @Roche about generative AI at the @appliedmldays AI & Molecular World track. 🧬🧑🏻‍💻 Many thanks for the excellent organization, @RebeccaNeeser @igashov @rneschneuing @damiano_sga @pschwllr @befcorreia #AMLDEPFL2024
3
1
33
1,192
Preprint of the presented talk: doi.org/10.26434/chemrxiv-20… and the paper of the presented poster: nature.com/articles/s41557-0…
3
130
ICML will be 🔥🔥🔥: Our workshop for ML for Life and Material Science: From Theory to Industry Applications is happening! Stay tuned for further announcements!
3
15
136
13,696
Looking for a computational scientist who's interested in a permanent position in Toronto. Physics background, and some QM & ML experience a must. If you're at the #APSMarchMeeting come talk to me after my talk!
2
43
1
107
18,471
Internship opportunity 🤓 - work with us on generative AI for antibody design Apply here: careers.roche.com/global/en/…
2
24
131
17,051
After 1.5 yeas our review “Integrating #QSAR modelling and deep learning in #drugdiscovery: the emergence of deep QSAR” in @NatRevDrugDisc is out! Great team effort with Alex Tropsha, Alexandre Varnek, Gisbert Schneider, and Art Cherkasov #compchem nature.com/articles/s41573-0… – at Pittsburgh, PA
2
23
1
114
11,149
Q: How to alter your training set selection if you only care about 1 prediction? A: Bias by similarity and train on the fly (basically kNN on steroids 🤩). Reaching chemical accuracy for a given query in QM9 this way implies 2 orders of magnitude less data than random training set selection. We also show how to derive density and dimensionality in chemical compound space. Very proud of our similarity based #machinelearning work with @Dom1Lemm and @ferchault #COMPCHEM
Machine Learning: Science and Technology@MLSTjournal6 Dec 2023
Great new work by @Dom1Lemm @ferchault @ProfvLilienfeld @UofTEngineering @acceleration_c @VectorInst @univienna @ml_tuberlin - 'Improved decision making with similarity based #machinelearning: applications in #chemistry' - iopscience.iop.org/article/1… #compchem #quantum #AI #materials
2
8
45
12,514
SAVE THE DATE! Chemical Compound Space Conference May 21-24 2024 in Heidelberg! Amazing speakers, posters, and science! Registration will open soon! More info here: ccsc2024.github.io/ Organized by @grynova_CCC @tristanbereau @raghurama123, Carolin Müller, Stefan Chmiela, and yours truly. #chemicalspace #machinelearning
6
1
53
12,025
A reaction screening approach combining HTP-experimentation with computational GNN to identify suitable substrates for late-stage C–H alkylation via Minisci-type chemistry, reported by @Roche, @ETH, @LMU_Muenchen, @DavidBKonrad, @atzkenneth, @ETH_DCHAB nature.com/articles/s42004-0…
3
14
3,313
Load more