iiinova — where your digital self keeps evolving

this is most of all true for the actual trainees. protecting them can inadvertently block them from opportunities they want. don't be that guy

020

overall, way fewer people are arguing against our policy as change seems inevitable now

A few quick bullets: protecting trainees from career risk is a common argument

100

Why careerism can be bad for science training

I spend a lot of time talking to potential grantees about our open science policy. Some field notes here: open.substack.com/pub/seemay/p...

open.substack.com

Some field notes from scholarly publishing discussions with scientists

173

Seemay Chou @seemaychou.bsky.social · 13/08/2025

A note in here for other science philanthropists… we have a chance rn to help scientists rethink systems, technical & operational. They’re game to take useful risks, let’s help them. Coming soon: more open science research, experimenting outside journals… seemay.substack.com/p/from-syste...

seemay.substack.com

061

It will certainly be important to include more modalities (and scientists) beyond X-ray, which we hope to do in the future! Lots of fun work to be done to ensure interoperability in that way

010

Stephanie Wankowicz @stephanieaw.bsky.social · 12/08/2025

Haha I love that, thanks for actually clicking the link 😂

000

Reposted by Seemay Chou

We are thrilled to be part of Project Diffuse, building infrastructure for the dynamic future of structural biology. We are helping to lead the modeling and encoding efforts of this project, including designing infrastructure that allows AI to learn from the full complexity of experimental data.

293

Reposted by Seemay Chou

Prism Science @prismscienceorg.bsky.social · 12/08/2025

Announcing the Diffuse Project! We're unlocking protein dynamics through diffuse X-ray scattering - the overlooked signal that could revolutionize how we understand protein motion. seemay.substack.com/p/from-syste...

seemay.substack.com

From systems operators to systems architects

Going up a level from data generation to think about the data systems we design and embed

196

Doing this will require co-iterating across an entire data system, not just refining one part. And we're experimenting with how to share all of this (our real-time thinking, pivots, tools, data) openly and quickly all outside of traditional journals

Tldr, we're focusing on diffuse scattering, which is the messier background signal in X-ray diffraction images. These data are often tossed because they're complex, but this complexity is where so much of the interesting biology sits

130

Our take: more of the same data won't get us there. We're bringing together top experts across almost every node of the X-ray pipeline to generate the right kind of data to model ensembles of different protein conformations. Think: pictures to movies

220

We're dreaming about the next frontier in structural biology: protein motion. Understanding how proteins move will be the next big unlock for predicting function

130

What’s the next PDB? Maybe it’s …the PDB. We just launched @diffuseproject.bsky.social a structural biology initiative exploring protein motion & rethinking how we generate and use experimental data. Quick 🧵👇 www.diffuse.science

diffuse.science

The Diffuse Project

45315

YouTube video by Arcadia Science

Most importantly, first in our new series of organismal spotlight shorts (open source illustrations coming soon!). Sea squirts! www.youtube.com/watch?v=-E9T...

youtube.com

Meet Ciona intestinalis

130

Hoped others would jump on it, but wasn't super straightforward so we ran with some predictions ourselves. Ultimately collaborating with sea squirt and choano experts to scope some pilots (biggest filter was finding people willing to publish openly/rapidly but we prevailed!)

zoogle.arcadiascience.com

We recently released the Zoogle dataset, which uses our workflows and publicly avail data to quantify how well different organisms model specific human genes/diseases. zoogle.arcadiascience.com/about

Zoogle | Arcadia Organism Selection Portal

Search for the best predicted organismal models for human genes.

121

After many yrs building internally @arcadiascience.com we're now able to partner more with others on cool organisms. You won't hurt my feelings if you don't read my post, but don't miss out on our first organismal spotlight animation featuring sea squirts -- just skip to the bottom! 🧵

1183

Props to those taking the first steps. Those first steps are scary, messy, and the only guarantees I can offer is that now we have a chance at something better. More announcements in coming months.

020

In the last 48h since announcing our policy, 4 academic PIs decided that they can indeed take this risk w us. They aren't the glam journal "elite" but they're damn good scientists. They're on the more junior side, and that's fucking awesome because they're our future.

#2 is actually a made-up concern in at least 50% of the cases that I help people work through. When you start talking to the real people involved, students are often hungry to try a different way but don't think their PIs will "let" them.

#1 makes sense because they actually think they have the most to lose since their success was built on the existing system. Why would they give up a winning hand even if it's in a broken game?

The "elite" scientists who publish in glam journals are actually waiting for the rest of you. When I talk to them, they either say: 1) Happy to do it once others do, and/or, 2) I can't do it because I have to protect so many trainees.

Seemay Chou @seemaychou.bsky.social · 03/06/2025

A v common suggestion I get re: publishing work outside journals is "if you just get some elite scientists to do it, everyone else will do it." This is an instinct that may seem sound, but here's what I've found irl:

140

I know this is a hard conversation to have, but it's really important that scientists lead it. Particularly at this moment. We might be able to really change things for the better, especially with all the technologies at our fingertips today.

251

Seemay Chou @seemaychou.bsky.social · 03/06/2025

We're excited to take this step, and to talk to other folks who are interested in experimenting. If you have something you're excited to try, reply here -- and we might even fund it. asterainstitute.substack.com/p/scientific...

asterainstitute.substack.com

Scientific Publishing: Enough is Enough

Why we're no longer funding journal publications

3289

Seemay Chou @seemaychou.bsky.social · 03/06/2025

We did something radical that shouldn’t feel radical: we stopped funding science built for journals so that we can reimagine scientific publishing. First @arcadiascience.com. Now at @asterainstitute.bsky.social.

25312

Seemay Chou @seemaychou.bsky.social · 25/04/2025

We @navigation.org are looking for proposals on how to move scientific publishing into the future. If you think we can do better, stop complaining and come build by applying AI towards open science os.nav.fund/ai-for-os/

os.nav.fund

AI for Open Science - Request for Proposals

010

Seemay Chou @seemaychou.bsky.social · 12/02/2025

Zoogle! Our new tool from @arcadiascience for picking research organisms. Even if you’re not a scientist, you should care that we spend >$50B/yr on clinical trials based on organismal models (>90% failure rate). Read more about Zoogle here: arcadiascience.com/blog/zoogle

arcadiascience.com

Introducing Zoogle: An Organismal Search Tool

Arcadia Science is pushing the boundaries of open science and innovating at every step in the research, development, and commercialization process.

0114

We want to hear from actual scientists about what the bottlenecks and specific opportunities are. As a microbiologist myself, I'm eager for this conversation and am invested in seeing it happen... by literally investing $ into it. Please help us do this in the right way.

000

This is going to be hard. Most functions are v complex. I don't think many are thinking realistically about how we predict causal basis without breaking the bank. Datasets need to public/structured, tools need to be open/maintained, problems need to be more defined beyond "bio foundation model."

100

And not just for microbes, as a window into eukaryotes too. Microbes not only have features that are conserved across the tree of life but also have honed extraordinary tools for studying other organisms they infect. We've learned a ton about basic cell biology from our microbial friends.

100

Request for Information - Astera Institute

Today we're sharing a new data initiative @asterainstitute.bsky.social near and dear to my microbiologist heart. W all the chatter about predictive models, what would it actually take to generate data that are broadly useful/interoperable for predicting functions? astera.org/request-for-...

astera.org

Data Requirements for Microbial Machine Learning Models What microbes can do Microbes have fundamentally shaped our bodies, our environment, and our understanding of biology. As a window into living s...

151

Seemay Chou @seemaychou.bsky.social · 24/12/2024

I’ll forgive @wilbanks.bsky.social for dropping the P-word (just this once) — if you’re curious about what kinds of data initiatives we’re gettin after @asterainstitute.bsky.social give his talk a listen. More deets on some big projects coming soon 👀 youtu.be/4Xnlf-sI0DM

youtu.be

Building scalable public data sets for scientific innovation

YouTube video by Roots of Progress Institute

161

Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...

Here's another cool example from Matt Daugherty's group on how tracing evolutionary signal gives us insights into how bio inventions like vertebrate eyes were "engineered" in the wild (my only complaint is that there is no preprint for this Matt 😉). www.pnas.org/doi/10.1073/...

pnas.org

PNAS

Picking the right research organisms/models in a data-driven way is a huge problem that will require all of us to fix -- no one denies this is a problem, no matter who you talk to (academia, pharma, etc). Yet we pour BILLIONS each year into clinical trials based on faulty animal models.

030

And an experimental example of this in action research.arcadiascience.com/pub/result-c...

See more here for examples of how you can take advantage of our work, on picking the right organisms research.arcadiascience.com/pub/idea-sel...

131