@sokrypton

Scientist, Assistant Professor @MITBiology, #FirstGen, ProteinBERTologist, 🇺🇦 No Human is illegal. Moving to: https://nitter.cf/t.co/sow6IRD3jj

 Cambridge, MA
solab.org
Joined December 2014
  • Tweets4.1K
  • Following3.8K
  • Followers19.3K
756Media
I'm excited to share that I'll be joining @MITBiology as an Asst Prof. in Jan 2024! Come join us! 🤓🧪🖥️🧬
170
152
10
2,004
223,044
Sergey Ovchinnikov retweeted
Garyk Brixi
@garykbrixi
3h
Genomes are full of dark matter of unknown functions. We present Minerva, a new method for discovery guided by genome language models. Minerva reveals the interactions hidden in non-coding DNA, pointing to hundreds of new putative RNAs and repetitive elements per bacterial genome. Minerva allows us to find and study elements invisible to traditional methods at orders of magnitude greater scale than before.
7
66
7
250
12,263
Sergey Ovchinnikov retweeted
Martin Pacesa@MartinPacesa
Sep 21
ʙɪɴᴅᴄʀᴀꜰᴛ2 is out, and we're not waiting for the paper. The full code drops today, free for academic and industry use. We're releasing it early so you can start designing right now, and bring its full power to the current Adaptyv competition. github.com/PacesaLab/BindCra…
26
317
23
1,428
136,476
Sergey Ovchinnikov retweeted
Ah-Ram Kim
@ahramkim1128
Sep 20
New in LIVIA: LIpDockQ and LIpDockQ2. pDockQ (Bryant et al. 2022) and pDockQ2 (Zhu et al. 2023) score full-length predictions, not the confident interface. On PPIs with IDRs or multi-domain proteins that dilutes the score like ipTM does, and neither filters by PAE. 🧵
Ah-Ram Kim
@ahramkim1128
May 13
🚀 New tool out! LIVIA (Local Interaction Visualization and Analysis) — a browser-based tool for assessing and visualizing predicted protein-protein interactions. Drop in a prediction from AlphaFold-Multimer, AF3, ColabFold, Boltz-1/2, Chai-1, or OpenFold3 (ZIP or folder, auto-detected) and LIVIA answers the two questions you actually care about: ▸ Do these proteins interact? ▸ Which residues form the interface? What you get: ▸ Interface confidence scores — iLIS (our local metric), ipSAE, actifpTM, ipTM ▸ Interaction interface heatmaps (PAE, LIS, cLIS) ▸ Sequence viewer + linear & circular contact maps highlighting Local Interaction Residues (LIR) and contact LIR (cLIR) ▸ Embedded Mol* 3D viewer ▸ Downloadable ChimeraX & PyMOL scripts Also fetches dimers directly from the AlphaFold Database — adding the interface annotations AFDB doesn't provide. Everything runs locally in your browser. No install, no upload. LIVIA started as a personal tool — I built it with Claude Code and used it to make every structure figure in our FlyPredictome preprint (Kim et al., 2026). (Claude Code is truly insane...) Along the way I realized it could be useful for others, too. If you have a favorite color palette for structure visualization, please let me know — happy to add it as a preset 🎨 🔗 LIVIA tool: flyark.github.io/LIVIA 🐍 iLIS / batch CLI: github.com/flyark/AFM-LIS 📄 LIVIA preprint: biorxiv.org/content/10.64898… 📄 FlyPredictome preprint: biorxiv.org/content/10.64898…
1
5
35
3,946
Should AI be credited as co-author on git commits?
47%Never, it's not liable
30%Always, credit matters
24%Only for real work
275 votes • Final results
5
2
3,114
ColabFold 1.6.3 is out! 2.5x faster, pip-installable, ipSAE+pDockQ2 scores. Thanks Choonghwan Lee, Marielle Russo, Gyuri Kim, Martin Steinegger, Milot Mirdita 🐍 pip install colabfold[alphafold] (1/3)
1
111
4
509
20,772
ColabFold2 Sneak Peak - integrates: alphafold3, openfold3, intellifold2, protenix2, boltz2, opendde, rosettafold3, chai1, esmfold2, openbind0 (thx Julia Buhmann), fully in Jax! 🗒️ Colab Notebook colab.research.google.com/gi… 🐍 pip install gist.github.com/sokrypton/75… (2/3)
4
43
160
7,048
Also a highly experimental port of these models into WebGPU, allowing direct prediction on your 💻: localfold.org WARNING: will drain battery🪫, overheat 🔥💻 & eat your data plan 📱 Started by Martin Steinegger as AF2 port to WebGPU martin-steinegger.github.io/… (3/3)
3
18
3
174
40,940
Private list of what I need next...
2
11
4,534
Sergey Ovchinnikov retweeted
Yunha Hwang
@Micro_Yunha
Sep 15
Microbial genomes have been extensively mined for new proteins, but the noncoding sequences between genes remain largely unexplored despite encoding important regulatory and functional information. In our new work, we asked whether gLM2 could help systematically discover these intergenic features. tatta.bio/intergenic-sae
3
39
3
204
11,421
Finally a more intuitive way to learn pLDDT/pAE? 😎 sokrypton.github.io/protein_… (Character idea from @HannesStaerk & Alex Waldherr)
55
279
44
1,372
161,882
Also credit to Gemini 3.8 and Claude Fable (now that their restrictions are relaxed and we can make protein fighting games) 🤪
1
1
47
4,939
Some updates: - Remote Multiplayer support fixed. - Adding support for custom PDB/AFDB inputs! (Initially implemented by @IanAndersonLOL )
1
2
21
1,452
Sergey Ovchinnikov retweeted
Aiden Kolodziej@aidenosinetrip1
Sep 8
My first, first-author work of my PhD is now out! pnas.org/doi/10.1073/pnas.26… Huge thanks to my incredible advisor @sokrypton and wonderful collaborator @MazAbulnaga for their support during this whole project!

Systematic discovery of circular permutations across the protein universe using CIRPIN | PNAS

Protein structure search has been revolutionized by deep learning methods that can rapidly search massive databases. However, current structure sea...

pnas.org
3
11
1
56
6,117
Sergey Ovchinnikov retweeted
Hannah Wayment-Steele
@HWaymentSteele
Aug 29
makeshift + ipympl = zoomable NMR spectra in Google colab 😎 catching the field up to the 2010's github.com/WaymentSteeleLab/…
2
12
1
72
5,056
Sergey Ovchinnikov retweeted
Claude
@claudeai
Aug 27
Starting today, 10,000 scientists across every field, from math to chemistry to physics and more, can get Claude through our new Claude Team plan for scientists. Standard seats are free, and premium seats with 5x usage limits are $15 per month, an 80% discount, for one year. Claude is becoming increasingly capable of scientific work, with recent progress on problems from advanced physics calculations to protein design. Alongside that progress, we've been investing in the research community: Claude Science launched in June, and our AI for Science program funds high-impact projects with free credits. Today's expansion builds on both. Principal investigators (or equivalent) at academic and nonprofit research institutions can sign up, then add the researchers in their group. Over the coming months, we plan to extend the program well beyond the initial 10,000 seats. Learn more: claude.com/programs/team-pla…
538
1,247
398
12,568
4,231,100
Sergey Ovchinnikov retweeted
Etowah Adams@etowah0
Aug 21
OpenBind intends to collect 10,000s of protein-ligand structures & affinities. To prioritize what we collect next, we need cofolding models trained on the latest data. Today we're releasing OpenBind-0 and 717 new ligand-bound structures.
4
60
6
206
22,499
Sergey Ovchinnikov retweeted
Hannah Wayment-Steele
@HWaymentSteele
Aug 21
Replying to @ginaelnesr
@ginaelnesr and I are excited to release makeshift -- a lightweight package to lower the barrier for using NMR dynamics data in ML. We hope this can pave the way for more NMR-based models like Dyna-1! Preprint: biorxiv.org/content/10.64898… Docs: makeshift-docs.readthedocs.i…
1
11
45
2,100
To me it seems silly everyone is trying to automate their tasks with LLMs. Wouldn't it make more sense to use LLMs to develop better workflows for their tasks that do not require calls to LLMs? 🤔
13
8
2
113
6,935
Looks like claude-protein-binder-design rediscovered our Protein Hunter protocol? 🤔@ChoYehlin
Yehlin Cho
@ChoYehlin
13 Oct 2025
Replying to @ChoYehlin
Protein Hunter: Starting from an all "X" sequence, we find that diffusion-based structure prediction models can hallucinate reasonable looking structures, which can be further improved through iterative sequence design and structure prediction, similar to AF2Cycler and LASErMPNN.
5
12
182
14,089
And implementing this idea is where most of their sucess is coming from... There are some references to Protein Hunter in the prompt, so I'm guessing that's where the agent got the idea from. But unfortunately, did not credit the idea... 🤷
2
5
58
3,553
additional analysis: nitter.cf/ChoYehlin/status/20902…
Yehlin Cho
@ChoYehlin
Aug 20
Replying to @ChoYehlin
Sorry about the missing axis labels😅 attaching the revised version
1
7
1,671
Load more