Happy to see MADRe published!!
MADRe: Strain-level Metagenomic Classification
academic.oup.com/gigascience/...
Great work @jlipovac.bsky.social

Following up on this - MADRe is now officially published 🎉

Very grateful for the guidance of @msikic.bsky.social @rvicedomini.bsky.social and Kresimir Krizanovic

🔗 academic.oup.com/gigascience/...

Myloasm, our long-read metagenome assembler, is now published! w/ @mgmarin.bsky.social and @lh3lh3.bsky.social

Very rewarding after > a year of development and countless hours thinking about assembly. Thanks to beta testers, Li lab, and reviewers who gave very helpful feedback.

rdcu.be/famFj

www.fer.unizg.hr/novosti/natj... under item 6 (in Croatian :) )

Hiring a PhD student in AI × RNA 🧬

• Predict, generate and optimise RNA 3D structure

📍 Zagreb (EU hub, great quality of life)
🌍 Team across Zagreb & Singapore (visit SG possible)

🧠 CS/math/physics + strong AI (not just prompting)
🧪 Real wet-lab validation of models
DM if interested.

Share 🙏 1/

Martin Schmitz makes history in the lab!
He successfully defended his PhD thesis, "Geometric Deep Learning for Diploid De Novo Genome Assembly," at the National University of Singapore, becoming the very first lab member to complete his doctorate there.

Congratulations, Martin!!

Science Has a Major Fraud Problem For decades, scientists were above reproach. Not any more. Joe Nocera investigates the murky world of fraudulent research, and the sleuths exposing dishonest science.

Science Has a Major Fraud Problem

'For decades, scientists were above reproach. Not any more. Joe Nocera investigates the murky world of fraudulent research, and the sleuths exposing dishonest science.'

@thefp.com

www.thefp.com/p/science-ha...

Great to have you back :). Your explanation of how Nanopolish works during a London calling many years ago gave me the idea for Racon

Thank you, Can!

This is an important step in the right direction. The methods are improving to start collectively thinking about how we should analyze raw nanopore signals for various new applications that are not possible without signal analysis. Congratulations to the authors!

Multi-context seeds enable fast and high-accuracy read mapping - Genome Biology A key step in sequence similarity search is to identify shared seeds between a query and a reference sequence. A well-known tradeoff is that longer seeds offer fast searches but reduce sensitivity in ...

1/ Our paper on Multi-Context Seeds is now out, with @tolyan.bsky.social spearheading the work and contributions from Nicolas and @marcelm.net. We introduce a new seeding concept that improves read alignment accuracy while maintaining speed.
link.springer.com/article/10.1...

Campolina: a deep neural framework for accurate segmentation of nanopore signals - Genome Biology Nanopore sequencing enables real-time, long-read analysis by processing raw signals as they are produced. A key step, segmentation of signals into events, is typically handled algorithmically, struggl...

Transformer-based AI has boosted
@nanoporetech.com
sequencing accuracy, but at a cost to portability due to GPU demands.
Our new work, spearheaded by
Sara Bakic, introduces Campolina link.springer.com/article/10.1... to improve nanopore signal segmentation for event-based mappers.

Thanks. Will look into it!

🧬 Hiring Postdocs at @astar_gis !

We need Computer Scientists and Computational Biologists to develop novel algorithms for de novo assembly of cancer genomes or to help us reconstruct them.

Experience in sequence alignment/assembly algorithms or assembly of complex genomes required

Please RT! 🔄

Join us this week for the RNA Salon at the Genome Institute of Singapore! Speakers from @boxiangliu.bsky.social (NUS) Dahai Luo and @msikic.bsky.social labs. Thanks for support by @rnasociety.bsky.social!

In accordance with Croatian laws, the funding is secured for up to six years.

Please share with strong candidates.

#AI #phd #RNA

The position offers close mentorship in a leading AI-for-RNA biology lab, access to substantial GPU resources, and full support to attend top international AI and ML conferences. There is also an opportunity to spend 1 to 2 years at
@astar-gis.bsky.social. 2/n

We are hiring a PhD student at the University of Zagreb, FER
for a project aimed at developing AI models to predict RNA structure and generate new RNA molecules.

We are looking for outstanding undergraduates in computer science, physics, or mathematics with demonstrated experience in AI. 1/n

7/ Interested?
📩 Send CV + short research plan
Please RT & share 🙏

#AI4Biology #AI4Genomics #ComputationalBiology #RNA #SingleCell #Hiring #PIPositions

6/ Who should apply?
Researchers with a strong computational background and a genuine interest in biology—from foundations to translation.

5/ Strong ecosystem 🤝
• Close collaboration with Institute of Molecular and Cell Biology (IMCB)
• Stable, long-term research funding in Singapore

4/ At scale ⚙️
• Automated experimental platforms
• Dedicated GPU clusters
• Access to A*STAR & National Supercomputing Centre Singapore compute resources

3/ What you’ll have access to 🧬
• DNA & direct RNA sequencing
• Single-cell & spatial transcriptomics
• Gene editing & RNA structure probing

2/ Our goal:
Build a place where AI scientists work tightly with experimentalists to create biologically verifiable models

1/ We’re hiring PIs in AI × Biology
The @astar-gis.bsky.social Genome Institute of Singapore (GIS) is expanding its AI & Computation domain and recruiting junior & senior PIs.

7/
Yes, many routine tasks will disappear.
But for hard biology, we will still need:

1️⃣ judgment and taste for what problems matter
2️⃣ genuinely new wet-lab techniques
3️⃣ new AI ideas inspired by biology itself

6/
This is the next battlefield—not just for agentic AI, but for biology itself.
Hard biological problems demand more than automation.

5/
Tang ends with the most provocative idea:
AI agents integrated with robotic wet-lab systems.

Closed-loop: hypothesis → experiment → data → model → next experiment.

4/
My take: many of these issues are solvable within ~3 years.
The more interesting question is economic & organizational:

Will this space be dominated by a few large players—or by open platforms where labs plug in their own protocols?

3/
Key obstacles remain:
• hallucinations that look correct but aren’t
• difficulty adapting agents to specific biological tasks

These are real, but largely technical problems.