Tessera
Folgen

Jonas Verhellen

@jonasverhellen.bsky.social

Theoretical physicist with a PhD in neuroscience. Postdoc in protein-protein interactions. Into art, science, and innovation. Currently: Copenhagen 🇩🇰 Previously: Oslo 🇳🇴, London 🇬🇧, Brussels 🇧🇪

37 Posts1696 Follower2097 Folgt
Marco Hernández@mhernandez.bsky.social · last yr.

Our team is looking for a visual journalist to report, design and code stories with us. You'll be able to learn from our skilled colleagues. I joined this team in 2021 and have continued to learn ever since. It's definitely a great work environment. Learn more about it or apply here: bit.ly/4kG33L6

The New York Times visual journalism story thumbnails
Ai2@ai2.bsky.social · last yr.

Introducing olmOCR, our open-source tool to extract clean plain text from PDFs! Built for scale, olmOCR handles many document types with high throughput. Run it on your own GPU for free—at over 3000 token/s, equivalent to $190 per million pages, or 1/32 the cost of GPT-4o!

Fernanda Duarte@fjduarte.bsky.social · 2y ago

1 week left to apply for this fully funded #compchem PhD opportunity part of the EU Marie Curie Doctoral Network, “CATALOOP” 🗓 Deadline: 31/01/2025 📥 How to Apply: submit your application via the university portal shorturl.at/nHyAo

Fernanda Duarte@fjduarte.bsky.social · 2y ago

We are seeking an outstanding #PhD candidate to join our team (www.duartegroupchem.org)! If you know potential candidates interested in combining #ML and #compchem reaction modelling for catalyst design, please share this opportunity with them! 🗓 Deadline: 31/01/25 📥 Application: shorturl.at/nHyAo

Jorge Bravo Abad@bravo-abad.bsky.social · 2y ago

Active-learning ML approach for catalytic free-energy calculations by Liu et al.: An MLP “committee” and constrained MD refine data, capturing reaction pathways 10^6× faster than ab initio. Their scheme ensures robust coverage of transitions and near-DFT precision. pubs.acs.org/doi/10.1021/...

CatFlow: An Automated Workflow for Training Machine Learning Potentials to Compute Free Energies in Dynamic Catalysis

Dynamic effects of catalysts play a crucial role in catalytic reactions, necessitating the incorporation of statistical sampling and understanding of the impact of dynamic structures in free energy calculations. However, the complexity of catalytic systems poses challenges in effectively exploring the vast configurational space effectively. In this work, we propose CatFlow, an automated workflow for training machine learning potentials (MLPs) to compute free energies of catalytic reactions. CatFlow combines constrained molecular dynamics (MD) simulation with concurrent training of MLPs and sequential calculation of free energies with well trained MLPs. By rapidly generating reliable MLPs, CatFlow facilitates rigorous free energy calculations, enabling the determination of the reaction profiles in an end-to-end manner. We showcased the capabilities of CatFlow by investigating the activation of O2 catalyzed by Pt clusters and demonstrated the effects of phase transition on the activities of the catalytic reaction. CatFlow offers an efficient and automated solution for studying the catalytic elementary reaction processes. It reduces the need for human intervention and provides researchers with a powerful tool to investigate free energies of dynamic catalysis.

pubs.acs.org

Jonas Verhellen@jonasverhellen.bsky.social · 2y ago

🔬 More #SciComm! 🔬 This week’s figure shows how severe mental disorders affect the entire body. 🧠 While these disorders are known for their impact on brain functioning (blue), patients are also disproportionately affected by physical diseases (red). Details 👇!

Severe Mental Disorders Affect The Entire Body. Severe mental disorders are characterised by their detrimental effect on brain functioning (shown in blue), but patients suffering from these diseases are also disproportionately affected by a range of other diseases (shown in red). For each cluster of diseases comorbid with schizophrenia or bipolar disorder, we highlight a representative organ and provide the 95% confidence interval mortality risk ratio in incident and prevalent schizophrenia cases versus the general population.
Jonas Verhellen@jonasverhellen.bsky.social · 2y ago

🔬 Hello, BlueSky! Time for some more #SciComm! 🔬 Last week’s (procrastination 🙈) figure illustrates the frequency spectrum of genetic risk factors for Schizophrenia: common variants (blue), protein truncating variants (red), and copy number variations (green). More 👇!

A figure illustrating the frequency spectrum of genetic risk factors for schizophrenia, with three categories represented by distinct colors: blue for common variants, red for protein-truncating variants, and green for copy number variations (CNVs). Each category is paired with a representative gene: C4A for common variants, SETD1A for protein-truncating variants, and NRXN1 for CNVs.
Jonas Verhellen@jonasverhellen.bsky.social · 2y ago

🔬 Hello, BlueSky! Time for some #SciComm! 🔬 This week’s figure illustrates the onset and progression of schizophrenia, integrating symptom intensity (blue line), environmental and genetic risk factors (red boxes), and key disease milestones (white boxes) as they unfold across age.

A graph illustrating the typical onset and progression of schizophrenia over time, with age on the x-axis and symptom intensity on the y-axis. A blue line represents symptom intensity, starting low in early childhood, increasing during adolescence, and peaking in early adulthood. Red boxes along the timeline highlight risk factors such as genetic predispositions and environmental triggers. White boxes mark key milestones in disease progression, including the prodromal phase, first psychotic episode, and chronic stages. A green-shaded area indicates a "window of opportunity" for early intervention, occurring before symptom intensity sharply increases.
Jonas Verhellen@jonasverhellen.bsky.social · 2y ago

🚀 Excited to announce that "Bayesian Illumination: Inference and Quality-Diversity Accelerate Generative Molecular Models" is now available! Key takeaway: Bayesian Illumination is 100x more effective than either genetic algorithms or deep generative models. 📊

Bild