←
marek mutwil 🇺🇦1,511 posts
marek mutwil 🇺🇦 avatar

marek mutwil 🇺🇦

@LabMutwil

Aim: improve gene function predictions of the plant kingdom to produce more climate change-resilient and nutritious crops.

◷ Joined October 2014▧ 313 Photos and videos
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Nov 16

Thrilled to share that [Your Name] has applied to do a PhD with us. The email was almost flawless: grateful, enthusiastic, and promising a complete application through the official portal, as instructed. Strong motivation. Impeccable manners. Identity: unknown. #Academia

Post attachment
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Nov 16

We’re hiring 4 PhD students at the University of Copenhagen (start 2027): LLM-built knowledge graphs for gene function, single-cell stress networks in crops, and 2 MSCA positions on plant metabolic pathways. Coding not required at entry. Deadline 31 Jan 2027.

marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Nov 13

5.6M genes from 149 plants in one aligned map 🌱 Aligned with ESCAPE, a non-rigid method from our autoresearch loop. Flowering plants merge into one shared core, while ferns, lycophytes & mosses keep their own islands. Compare ORBIT / THESEUS / ESCAPE: claude.ai/artifact/CkW8Uy8Gh…

Post attachment
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Oct 16

3/4 It could be competitors, AI data collectors or someone else. We cannot identify who is behind it. The impact is clear: during the worst period sampled, almost one in four requests to CoNekT failed, making the database harder for researchers to use.

marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Oct 16

2/4 The traffic suggests considerable effort to collect our gene, sequence and co-expression data: spreading requests across many internet addresses and disguising automated software as ordinary browsers.

marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Oct 16

1/4 Our CoNekT and PEO databases are being heavily scraped. From 7–15 October, CoNekT’s main site recorded 14.3 million requests. On 14 October, about 88% of 3.7 million requests matched two automated scraping patterns.

marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Sep 13

🌱 3 PhD positions for 2027 in computational plant biology & AI at @UCPH_Research! Mutwil Lab seeks curious minds who question data and see the bigger picture. Prior coding experience optional; willingness to learn essential. Details & initial enquiries: publish.obsidian.md/mutwilla…

marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Sep 3

One example from the evaluation process: Eight external LLM agents played 44 fights using fixed seeds. Seven defeated the first boss. This is an informal development experiment, not academic research or a measurement of human player difficulty. 4/5

Post attachment
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Sep 3

I have used coding agents, PixelLab for visual development, and OpenRouter models for automated playtesting. The interesting question is not whether AI can generate output. It is whether that output remains useful, correct and coherent across a large project. 3/5

Post attachment
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Sep 3

The game is also something I genuinely want to play. It combines 1990s pixel art, difficult roguelikes, hard science fiction and deckbuilders with complicated synergies. Enemy ships carry their actions along a shared timeline and can be destroyed before those actions occur. 2/5

Post attachment
marek mutwil 🇺🇦 avatar
marek mutwil 🇺🇦@LabMutwil·Sep 3

A personal project outside my usual computational biology work: I have been building The Long Ascent, a turn-based deckbuilding game, to explore how far one person can take a complex project using current AI tools and a data-science mindset. mutwil.github.io/ 1/5

Post attachment