ChemRxiv (@ChemRxiv) header
ChemRxiv (@ChemRxiv) profile picture

ChemRxiv is the open preprint server for the global chemistry community. Contact us at [email protected] for assistance.

Joined January 2017
Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)Photo by ChemRxiv (@ChemRxiv)
We're now on BlueSky! Come find us at go.acs.org/cKN
2
1
7
12,429
ChemRxiv retweeted
Check out our latest preprint on @ChemRxiv (dx.doi.org/10.26434/chemrxiv…), where we report some of the brightest red #TADF #LECs to date. This work was done in collaboration with the group of Ludvig Edman. @StAndrewsChem @StAndrewsOsc
1
8
50
2,981
My colleagues from @inductive_bio just published a preprint on @ChemRxiv that they will present at @NeurIPSConf @AI4D3 in Sydney. It’s the latest piece of work from Inductive on a theme that’s near and dear to our hearts: how to rigorously evaluate AI models for molecular property prediction. Good models are born out of good evals, but it’s surprisingly hard to evaluate molecular property prediction models. Our very first blog post from 2023 addressed one aspect of this: benchmark datasets that existed at the time did not accurately reflect the challenges of real drug programs. A lot has changed in the last three years. First, the number of papers with new molecular representations has exploded, many of which claim state-of-the-art performance. As a researcher in the field, this is both exciting and overwhelming – it can be hard to figure out which representations to use in practice. Luckily, something else has changed in the last three years: publicly available benchmarks for molecular property prediction have gotten much better, thanks in large part to the work of groups like OpenADMET. The work in this paper builds on previous work from researchers at the AGH University of Krakow to systematically evaluate 48 molecular representations across 17 ADMET endpoints from the first three OpenADMET blind challenges. Along the way, it improves the statistical methodology for comparing a large number of models across a large number of datasets, incorporating the magnitude of performance differences instead of just the direction and correcting an assumption from earlier work that led to overestimating confidence. The end result showcases the exciting progress the field has made recently: the paper finds that 29 of 47 recent representations outperform ECFP in terms of latent ability, with Monroe and Atom-JEPA as clear standouts. My hope is that rigorous evaluations like this, paired with the proliferation of high-quality benchmark datasets will continue the acceleration we are seeing in molecular property prediction. Kudos to Ne Dassanayake (our summer intern from the @cwcoley lab), Drew McNutt, and @alex_s_rich for awesome work. Here's the link to the full article: chemrxiv.org/doi/full/10.264…
4
6
55
3,270
ChemRxiv retweeted
🧩 New building blocks for peptide simulations! 👉 doi.org/10.26434/chemrxiv.15… Our latest @ChemRxiv preprint introduces an MB-nrg library of #datadriven #manybody potentials for all 20 standard amino acids, a step toward CCSD(T)-level simulations of peptides! We combine #machinelearned n-body terms trained on DLPNO-CCSD(T) fragment data with physics-based electrostatics, polarization, and dispersion. These reusable building blocks let us assemble potentials for complete dipeptides without using whole-dipeptide energies for training. 😎 This is the first iteration of the potentials. We’ll keep improving and expanding them as we develop larger training sets of n-body interactions, including additional nonbonded and higher-body terms. 🏄‍♀️ Great work by Xinmao, Ethan, and Jiwon! 🙌 Interested in performing #datadriven #manybody simulations? Check out #MBX: ✅ mbxsimulations.com ✅ github.com/paesanilab/MBX ✅groups.google.com/g/mbx-user… Big thanks to @AFOSR for supporting the development of these MB-nrg potentials for amino acids and to @NSF for supporting their implementation in #MBX! We are also grateful to #HPCMP and @ACCESSforCI for providing computational resources! 🙏 @UCSanDiego @UCSDPhySci @UCSDChem @HDSIUCSD @SDSC_UCSD
1
4
34
2,219
ChemRxiv retweeted
🚨Our latest preprint is now out on @ChemRxiv !🚨 For over 40 years, this reactivity went unnoticed: at room temperature, silyl- and germyl-thioketones quietly undo the aromaticity of their own benzene rings. No light, no redox, no transition metals. chemrxiv.org/doi/full/10.264…
23
1
103
5,185
ChemRxiv retweeted
Check out our latest @ChemRxiv (dx.doi.org/10.26434/chemrxiv…), where we show that overlap b/w the #photoluminescence & #absorption spectra drives singlet exciton diffusion & predicts the magnitude of efficiency roll-off in #tadf #oleds. This was in collab/w the group of Ifor Samuel.
4
1
21
1,915
🪜Hetero-[2]-Ladderanes: Our work on the first diastereoselective synthesis of oxa-aza-[2]-ladderanes, a potential 3D mimic of 2,3-disubstituted pyridines, is now on @ChemRxiv. Thank you, @ANRFIndia and @iiscbangalore, for financial support. bit.ly/4AuNuNR
10
21
252
8,243
Ladderanomania! In @ChemRxiv today: bit.ly/4eqyg3R
2
13
1
107
6,624
ChemRxiv retweeted
Have a look at the preprint of @clment_tanguy and Matt Wodrich @ChemRxiv describing the generation and use of azirinyl radicals and azirinium cations starting from the same redox active ester precursors under visible light irradiation: chemrxiv.org/doi/full/10.264…
4
1
92
4,699
One molecule, many electronic descriptions. Our preprint on @ChemRxiv introduces Constructive Chemical Space (CCS)—a formal framework linking molecular identity, electron allocation, and stepwise construction across the entire molecular chemical space! chemrxiv.org/doi/full/10.264…
3
12
1,349
ChemRxiv retweeted
Poly(pentafulvalene) is a more stable polyene with a narrower band gap! I'm excited to share our new preprint on @ChemRxiv ! Using oligomers, we demonstrate how five-membered rings can narrow the bandgap. And don't miss the reversible 8-electron reduction! chemrxiv.org/doi/full/10.264…
1
9
44
3,420
Spherical SiliaCat Pd(0) selectively catalyzes the hydrogenolysis of N-Cbz-protected amino acids and dipeptides under ultramild conditions. New achievement in green chemistry with eminent colleagues in Québec led by F. Béland openly readable @ChemRxiv: shorturl.ly/Zt_ob
🤖 Made with AI
2
1
3
787
KFB Lab Alert: Our new screening workflow for VHPO reaction development (Haloplate) is published at ACS Bio & Med Chem Au! Congratulations to @sp3intermediate for leading this work alongside of Sergio, @HannahSlocumb5, and @jbacsa1! pubs.acs.org/abmcb8/article/…
KFB Lab Alert: One Biocatalyst, LOTS of N-Heterocycle Halogenations! Our latest work on the development of an efficient screening workflow for VHPO halogenation is now online @ChemRxiv. Huge congratulations to @sp3intermediate for leading this work with an awesome team of Sergio, @HannahSlocumb5, and @jbacsa1! #Haloplates chemrxiv.org/doi/full/10.264…
5
20
1,585
ChemRxiv retweeted
New @ChemRxiv, study on using hydroxyethyl acrylate to turn epoxy-amine thermosets into covalent adaptable networks. doi.org/10.26434/chemrxiv.15…
1
2
702
📢 Excited to share our recent work where we demonstrate the metal-free, photoinduced divergent activation of robust gases (N₂O and H₂) under ambient conditions! Now out @ChemRxiv #maingroupchemistry #photochemistry #smallmolculeactivation 📄 Read here: doi.org/10.26434/chemrxiv.15…
1
4
66
3,631
One main-group platform, two inert gases! ⚛️ Excited to share our latest work on the light-driven activation of both N₂O and H₂ using silylphosphanes at ambient conditions! Read it here @ChemRxiv doi.org/10.26434/chemrxiv.15… #maingroupchemistry #smallmolecules #photochemistry
1
21
1,684
With no one watching, a CRISPR diagnostic test can malfunction and still read "negative"! Our new Perspective is out on @ChemRxiv today: "AI Agents for CRISPR Diagnostics: From Automation to Bounded Autonomy" chemrxiv.org/doi/full/10.264… #CRISPR #AIAgent #Diagnostics #LabAutomation
1
501
ChemRxiv retweeted
Check out our latest in @ChemRxiv (dx.doi.org/10.26434/chemrxiv…), a collaboration with the group of Girish Lakhwani, looking into the photophysics of #tadf materials. @StAndrewsChem @StAndrewsOsc @TADFSolutionsDF
2
7
1,243
New BORONPro preprint from the @The_Ye_Lab, Lin, Finze and @LabBraunschweig groups! @ICBWuerzburg
A Crystalline Aminoborane Radical Defines a New Stabilization Strategy and Enables Dihydrogen Activation and Transformation ( @ChemRxiv ): chemrxiv.org/doi/full/10.264… ( @The_Ye_Lab ; @LabBraunschweig ; @SFBBoronPRO ).
2
13
1,187