#diffdock
🧬 Testing beta-lactamase resistance with AlphaFold + DiffDock + GROMACS! Watch clavulanic acid bind TEM-5, CTX-M-15, KPC-2, and get rejected by TEM-30. Simulation confirms biology! 🔬💊 #gromacs #diffdock #chimerax #moleculardynamic #alphafold
Assessing TEM, CTX-M, and KPC2 With Molecular Docking and Molecular Dynamic Simulation | Everyday Is A School Day
🧬 Testing beta-lactamase resistance with AlphaFold + DiffDock + GROMACS! Watch clavulanic acid bind TEM-5, CTX-M-15, KPC-2, and get rejected by TEM-30. Simulation confirms biology! 🔬💊
www.kenkoonwong.com
February 22, 2026 at 3:49 AM
Not happy either! It would have been better to have our offer of help accepted by Gabriele and the rest of the DiffDock authors. You, me, and @annclevesjain.bsky.social have close to 100 person-years of experience in this area. We offered our help for free!
December 9, 2024 at 1:36 AM
MIT의 AI 모델 DiffDock이 100초 만에 신약 enterololin의 작용 메커니즘을 예측하고, 연구진이 6개월 만에 검증에 성공했습니다. 기존 2년/200만 달러 소요 작업을 획기적으로 단축한 AI 기반 신약 개발 사례를 소개합니다.
AI가 과학자보다 먼저 답을 맞힌 날: 신약 개발 18개월→6개월 단축의 비결
MIT의 AI 모델 DiffDock이 100초 만에 신약 enterololin의 작용 메커니즘을 예측하고, 연구진이 6개월 만에 검증에 성공했습니다. 기존 2년/200만 달러 소요 작업을 획기적으로 단축한 AI 기반 신약 개발 사례를 소개합니다.
aisparkup.com
October 7, 2025 at 3:35 AM
DiffDock, a new diffusion-based ligand docking program, made s big splash earlier this year. But it’s apparently not all it’s claimed to be:
Computational Care
www.science.org
December 6, 2024 at 5:01 PM
😮
"And the preprint highlights a major problem behind the DiffDock team's approach. As a training set, they used roughly 98% of the known small-molecule-bound structures in the literature, and reserved the other 2% to evaluate the program's performance as a test set."
DiffDock, a new diffusion-based ligand docking program, made s big splash earlier this year. But it’s apparently not all it’s claimed to be:
Computational Care
www.science.org
December 6, 2024 at 8:22 PM


DiffDock-Glide: a hybrid physics-based and data-driven approach to molecular docking

https://www.biorxiv.org/content/10.1101/2025.06.02.657461v1
June 9, 2025 at 1:24 PM
While dockers are keep docking (now with diffusion and AI😀), Pat Walters and Ajay Jain show sobering assessment of ‘perceived’ accuracy of such methods! Diffdock shots fired 🔥 The conclusion is totally worth reading in full! #chemsky #compchemsky
arxiv.org/abs/2412.02889
Deep-Learning Based Docking Methods: Fair Comparisons to Conventional Docking Workflows
The diffusion learning method, DiffDock, for docking small-molecule ligands into protein binding sites was recently introduced. Results included comparisons to more conventional docking approaches, wi...
arxiv.org
December 6, 2024 at 5:33 AM
Foundation models like AlphaFold2 and DiffDock 2.0 are accelerating drug discovery, reducing molecular screening time by up to 6.2x ⏱️.

Platforms like NVIDIA’s BioNeMo optimize these processes, making in silico screening faster and more scalable than ever.
April 21, 2025 at 8:36 PM
imo not a ideas vs solutions issue but a arrogant ML researchers vs. the rest issue. they could've carefully reported their results, they didn't. i know of several people that wasted a substantial amount of time confirming diffdock is useless for any CADD task. good this paper is here to warn others
December 7, 2024 at 9:33 PM
Recent dispute over DiffDock was quite mesmerizing: DiffDock original preprint... Jain's validation preprint...Lowe's blog...Gabriele's response on LinkedIn...Pat's response to Gabriele's. My understanding is disciplined and standardized benchmarking is the key. A succinct paper that might help bit.
Data standards in drug discovery: A long way to go - PubMed
Each year, millions to trillions of data points are generated to evaluate the response of chemicals and biologicals to human cells in vitro and in vivo using various technologies and endpoints. Despit...
pubmed.ncbi.nlm.nih.gov
December 9, 2024 at 3:10 AM
⚛️ cuEquivariance ⚛️ is something I've been waiting for forever. Thank you Mario Geiger and team!

Curious to see where the "equivariance is just too slow" debate goes 👀

github.com/nvidia/cuequ...
GitHub - NVIDIA/cuEquivariance: cuEquivariance is a math library that is a collective of low-level primitives and tensor ops to accelerate widely-used models, like DiffDock, MACE, Allegro and NEQUIP, ...
cuEquivariance is a math library that is a collective of low-level primitives and tensor ops to accelerate widely-used models, like DiffDock, MACE, Allegro and NEQUIP, based on equivariant neural n...
github.com
November 21, 2024 at 9:47 PM
Speaking of which, we've got 60us MD simulations of these 50 ligands binding to two different 20 residue a-syn fragments, so if someone wants to help us take this data and retrain IDP versions of DiffDock or AutoDock- let us know!
January 31, 2025 at 7:59 PM
We find that, on average, AutoDock is in closer agreement with MD than DiffDock. DiffDock predictions seem to be primarily driven by aromatic stacking, and DiffDocked ensembles don't have many hydrogen bonds or charge contacts.
January 31, 2025 at 7:59 PM
Autodock Vina and Diffdock both produce the same ranking of the three ligands we see by NMR CSPs and MD. Righteous!

But do the ensembles and molecular interactions match MD? You betcha! Check out the similarity of AutoDock Vina and MD for this cluster of a-syn-C-term.
January 31, 2025 at 7:59 PM
With AutoDock Vina, we dock on each residue of each MD frame, restricting the ligand search space to a volume around that residue. We take the best scoring pose as the prediction for that frame.

With DiffDock we just point and shoot once at each frame and let AI do its diffusion magic
January 31, 2025 at 7:59 PM
One note about the DiffDock *preprint* fracas: it seems like two of the main issues at hand

i) experienced researchers insisting on addition controls to justify claims

ii) objections to overly combative language and potentially misleading comparisons

are things that peer reviewed journals...
December 10, 2024 at 5:49 PM
when my exams are over I should write something contextualizing DiffDock within the long history of software people ignoring domain experts
December 9, 2024 at 8:50 PM
Boltz-1の開発グループはDiffdockと同じグループなんだけど、なんでかDiffdockと性能比較していないね あとよく言われるAutodockとも。
するまでもなく性能良くなっているんだろうけれど、論文の査読者的には「なぜ比較していない?」って言うのかな
August 19, 2025 at 6:28 AM
The DiffDock authors made a response to the recent critiques. Nothing but fair to share them here too!

Surely now the experts will have a technical discussion about it and push everyone forward.

As I have been trying to advocate, let's all breath a little longer and give things more time.
You may have seen a recent pre-print [1] from Jain et al. with strongly worded claims against the experimental results in our DiffDock paper [2]. We initially declined to respond as we saw that this preprint contained falsehoods, misleading comparisons, seemingly deliberate omissions, ...1/n
December 9, 2024 at 6:53 AM
Just to be clear: I don't want to imply that the DiffDock authors caused the hype. Just that it can happen, and in science we should mostly try to avoid.

Anyone can suffer from this, including ORCA devs.
December 6, 2024 at 7:36 AM
I'm not happy with the way this all went down. I have tremendous respect for Gabriele and the rest of the DiffDock authors. Their work has broken new ground and helped advance machine learning in drug discovery. If I had to do it over again, I'd do things differently.
December 9, 2024 at 1:17 AM