Tessera
Folgen

Nils Weskamp

@nilsweskamp.bsky.social

Teaching computers chemistry. Helping data sets with people problems. Sailing the sea. Opinions my own.

22 Posts207 Follower716 Folgt
Michael Bronstein@mmbronstein.bsky.social · last yr.

Join us in Vienna on Sep 8-10 with an incredible speaker lineup!

AITHYRA Research Institute of the Austrian Academy of Sciences@aithyra.bsky.social · last yr.

Please save-the-date for the #AITHYRA Symposium “AI for Life Science”, which will take place in Vienna at the House of Industry from 8-10 September 2025. Great speakers, interesting framework programme and beautiful location. Registration will open mid May. More info: www.oeaw.ac.at/aithyra/news...

Michael Bronstein@mmbronstein.bsky.social · last yr.

Next week we will be at ICLR in Singapore, presenting 8 papers in the main conference as well as in multiple workshops. For anyone interested, I will be glad to talk about PI positions at Aithyra (spoiler: we offer some of the best conditions in Europe)

Nils Weskamp@nilsweskamp.bsky.social · last yr.

Integrating AI/ML-based approaches with mechanistic modelling has lot of potential for molecule design in drug discovery. Happy to see this published! doi.org/10.1111/cts.... #ChemSky #MedSky

Predicting Pharmacokinetics in Rats Using Machine Learning: A Comparative Study Between Empirical, Compartmental, and PBPK‐Based Approaches

A successful drug needs to combine several properties including high potency and good pharmacokinetic (PK) properties to sustain efficacious plasma concentration over time. To estimate required doses...

doi.org

Phil Biggin@philbiggin.bsky.social · 2y ago

Very nice to see the work of Matt Warren and Isak Valsson just out with Charlotte Deane, Aniket Magarkar and @garrettmm.bsky.social where we show how augementation can dramatically improve ML affinity predictions to close to FEP accuracy. rdcu.be/d88Wm

Narrowing the gap between machine learning scoring functions and free energy perturbation using augmented data

Communications Chemistry - Machine learning offers great promise for fast and accurate binding affinity predictions, however, current models lack robust evaluation and fail on tasks encountered in...

rdcu.be

Nils Weskamp@nilsweskamp.bsky.social · 2y ago

"Drug discovery & development continues to be challenging, and despite a desire to find short cuts, guides, mnemonics, and simple predictive parameters as surrogates for clinical success, we must continue to critically assess how useful these methods are." www.nature.com/articles/s41... #ChemSky

Return to Flatland - Nature Reviews Chemistry

Fifteen years ago, an analysis of drug molecules moving through the stages of clinical development suggested that increased three-dimensional character was a marker of greater success. Now, we perform...

nature.com

Francesca Grisoni@fragrisoni.bsky.social · 2y ago

If you use generative #DeepLearning for molecule design, check out our latest work, where we perform a large scale analysis (~1 B designs!) and find ‘traps’, ‘treasures’ and ‘ways out’ in the jungle of generative drug discovery. 🌴 🐒 Paper: arxiv.org/abs/2501.05457 Code: github.com/molML/jungle...

The Jungle of Generative Drug Discovery: Traps, Treasures, and Ways Out

"How to evaluate de novo designs proposed by a generative model?" Despite the transformative potential of generative deep learning in drug discovery, this seemingly simple question has no clear answer...

arxiv.org

Nils Weskamp@nilsweskamp.bsky.social · 2y ago

To augment human intelligence with digital tools, we must understand what the human brain is great at and where it has limitations. This paper estimates that human behavior happens at 10 bits per second, so processing information at scale should be on the list: www.sciencedirect.com/science/arti...

The unbearable slowness of being: Why do we live at 10 bits/s?

This article is about the neural conundrum behind the slowness of human behavior. The information throughput of a human being is about 10 bits/s. In c…

sciencedirect.com

Nils Weskamp@nilsweskamp.bsky.social · 2y ago

Before you put your next presentation together, you should have a look at this classic paper. If you want the audience to remember what you said, it has to be a story. #Storytelling #ScienceCommunication link.springer.com/article/10.3...

Narrative stories as mediators for serial learning - Psychonomic Science

Subjects learned 12 serial lists of 10 nouns by one of two methods: a control method of normal study and rehearsal, or a narrative-chaining method, where S was instructed to construct a meaningful sto...

link.springer.com