Greg Medlock

Greg Medlock

@gregmedlock.bsky.social

VP Technology @ Vedanta Biosciences. Interested in repairing+controlling the human microbiome.

1.2K followers803 following183 posts

Fascinating story on the effects of C. difficile colonization in infants.

Joe Zackular@joeyzacks.bsky.social

Excited to share new work from my lab showing that early-life colonization with C. difficile remodels the developing infant gut, published today in @science.org! This study was co-led by two truly amazing scientists: @haidermanzer.bsky.social and Alexa Semon. www.science.org/doi/10.1126/...

Greg Medlock reposted

couple months ago, I was riding shotgun on a project that went sideways overall premise was good, it was aggressively spec'd, then it was fed to an agent, which completely mangled the intention in subtle ways this got me thinking about the missing spine for agent dev github.com/daniloc/cohe...

github.comGitHub - daniloc/coherence: The missing spine for agent-driven coding. Create the structure you need to reduce the cost of changes, verification and fixes.The missing spine for agent-driven coding. Create the structure you need to reduce the cost of changes, verification and fixes. - daniloc/coherence

AI shouldn't be doing review *for you*, but 1) if you've submitted a manuscript and it's not also preprinted and in the public domain already, that's careless, 2) current generation AI is better than the best humans at some aspects of peer review--it would be a disservice NOT to use it

Itai Yanai@itaiyanai.bsky.social

If you’re letting AI do a review of a manuscript for you, just know that (A) it’s super obvious to anyone reading it, (B) it’s dishonest - as you’re explicitly instructed not to do so, and (C) it’s careless - because you’re uploading someone’s confidential material to an AI app.

ICME abstracts are due this Wednesday, Aug 5! Please consider submitting your work.

Greg Medlock@gregmedlock.bsky.social

I'm excited to help organize #ICME2026 - join us to explore the latest advances in Microbiome Engineering! Our call for abstract is open and we are seeking sponsors. Please share, and reach out to me to learn more!

Greg Medlock reposted

In May, Vijeta and I had our wedding in Cambridge Commons with hundreds of you all. In June, her grandfather suddenly passed away. In July, her papa was diagnosed with stage IV cancer and a brain tumor. He is 58. He was at our wedding, already symptomatic, and none of us knew.

I'm excited to help organize #ICME2026 - join us to explore the latest advances in Microbiome Engineering! Our call for abstract is open and we are seeking sponsors. Please share, and reach out to me to learn more!

Greg Medlock reposted

“entire 2026 budget” doing some heavy journalistic lifting here. As a large public co, 2026 budgets were probably more or less final pre-Opus 4.5 (maybe even pre-4.1). Very curious what the actual $$ was!

Dare Obasanjo@carnage4life.bsky.social

Uber's COO recently voiced an uncomfortable truth that many in the tech industry are yet to admit. He said while Uber's engineers are consuming massive amounts of AI tokens, it is incredibly difficult to prove that this spending actually benefits the end user or improves the product proportionally

Just released an update of Medusa, a Python package for managing ensembles of genome-scale metabolic models (github.com/opencobra/Me...) This release incorporates improvements by Jeroen Gilis from Chalmers University, including speeding up ensemble generation from a set of SBML files >10X

github.comGitHub - opencobra/Medusa: Analysis of ensembles of metabolic network reconstructionsAnalysis of ensembles of metabolic network reconstructions - opencobra/Medusa

Michael, this reads as downright conspiratorial. Have you interacted in any way with NSF staff working on this? There’s extensive oversight. No/minimal human subject work is even likely to be funded. There is independent merit-based review. Where are you getting your info?

This post is unavailable.

Greg Medlock reposted

Save the date - join the 9th International Conference on Microbiome Engineering! 📍Durham, NC 🗓️ Dec 1-3, 2026 Are you obsessed with controlling microbiomes? Want to learn all the ways accomplishing this goal benefits society? This is the meeting for you! www.aiche.org/sbe/conferen...

aiche.org9th International Conference on Microbiome EngineeringThis conference brings together scientists across the quickly growing field of microbiome engineering. The conference integrates synthetic biology, systems biology, microbial ecology, and bioinformati...

The Anthropic team does lots of helpful posting to X, but unfortunately has not prioritized bluesky cross-posting (makes sense, X algo loves everything AI) @j4ck.xyz made this tool to mirror those accounts, and takes community requests. Boris @bcherny-x.bsky.social Thariq @trq212-x.bsky.social

jack@j4ck.xyz

I’m glad it’s useful, I’ve been trying to bridge useful Twitter accounts here for this very reason! There’s a list I made of all bots here → bluesky-mirrors.vercel.app

Bonus: with these mirrors, your feed doesn't get populated by anthropic's gracious replies to "tHeY NErFeD dA MoDEl" or "what do you mean I can't deploy Claude as a service in my app using my $100/month subscription?" engagement bait.

Trying to reduce my use of social media with algorithmic feeds, and found this very nice bot today. So much of Anthropic’s dev and media teams just post to X as a sole outlet. 2nd best is subscribing to new release notes for Claude code on github.

Anthropic {bot}@anthropicai.xmirror.bot

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.

Vedanta Biosciences Announces the Phase 3 RESTORATiVE303 Study of VE303 for the Prevention of Recurrent C. difficile Infection Will Continue as Planned Following Protocol-Specified Interim Analysis

This is wild, we eat so much yeast

Manuel Kleiner@manuelkleiner.bsky.social

I thought it was quite mind boggling to see that around 10% of the protein in a standard yeast based bread is actually yeast protein! Here our #metaproteomics analysis of microbial protein in fermented foods pubs.rsc.org/en/Content/A... Congratulations to first authors Laura Winkler and Ayesha Awan

Greg Medlock reposted

If you're a senior PhD student or recent graduate interested in doing work at the intersection of AI and microbial genomics and/or microbiomes, please feel free to reach out to discuss the possibility of applying together for an MIT Novo Nordisk Fellowship engineering.mit.edu/novo-nordisk

engineering.mit.eduNovo Nordisk FellowshipPotential areas of focus for postdoctoral fellows participating in the MIT-Novo Nordisk Postdoc Program include, but are not limited to, the following: Your complete application should include: In the...

CARB-X is opening a new funding round soon—sign up for their newsletter (link in replies) to get the announcement. CARB-X is one of few organizations making meaningful progress ensuring access to new antimicrobial drugs and supporting preventatives/diagnostics.

Tried Edison Analysis (not Kosmos) on two tasks and was blown away: 1. Uploaded 96 high-res NMR spectra of bacterial supernatants and asked for metabolite ID and quantification 2. Uploaded a bacterial genome and asked it to identify mobile genetic elements (more complex, but this is the gist)

Sam Rodriques@sgrodriques.bsky.social

Give it a try here: platform.edisonscientific.com And read our paper: edisonscientific.com/kosmos-report We have built this to accelerate science, and we think it's pretty neat. We are super excited to see what people think.

View full thread

For the genomic analysis, Edison performed similarly, but "connected the dots" within the analysis in a way it is hard to imagine a PhD student doing on their own 5 years ago.

Really encourage folks who had bad impressions of earlier LLMs & hallucinations to take a look at domain-specific, agentic systems like Edison, Claude Code, etc. Semantics aside--these things are just very useful now.

Tried Edison Analysis (not Kosmos) on two tasks and was blown away: 1. Uploaded 96 high-res NMR spectra of bacterial supernatants and asked for metabolite ID and quantification 2. Uploaded a bacterial genome and asked it to identify mobile genetic elements (more complex, but this is the gist)

Sam Rodriques@sgrodriques.bsky.social

Give it a try here: platform.edisonscientific.com And read our paper: edisonscientific.com/kosmos-report We have built this to accelerate science, and we think it's pretty neat. We are super excited to see what people think.

I knew what reasonable results should be and how long it would take a human. For the NMR task, Edison performed similarly to a senior PhD student that manually reviewed the entire spectra, and correctly identified things like negative controls without instruction.

Greg Medlock reposted

Antimicrobial resistance (AMR) is one of the greatest threats to public health, with the potential to undo decades of medical progress. This #WAAW2025, head to our LinkedIn to hear from world-leading experts on how we can tackle AMR and protect the future of medicine: t.ly/3Oxk1

t.ly#worldamrawarenessweek #waaw2025 | LifeArcAntimicrobial resistance (AMR) is one of the greatest threats to public health, with the potential to reverse decades of medical progress. Despite the severity of the problem, the pipeline for new dr...