I'm excited to share our new preprint: PocketVina — a fast, scalable, and accurate multi-pocket molecular docking method.

Docking remains essential in early-stage drug discovery, but recent deep learning–based approaches still face limitations in generating...

Thread - (1/n)

Portrait photo with text overlay: The image shows a man in a white button-down shirt standing outdoors with arms crossed, facing the camera and smiling slightly. The background is filled with greenery and blurred buildings. Overlaid text in white sans-serif font appears in the lower left corner, reading “ALTUNA AKALIN,” with “AKALIN” in bold uppercase, all inside two light blue text boxes.

AI can’t replace your doctor – but it might help you find the right one.

New study by @al2na.bsky.social from #mdcBerlin:
✔️ 97% accuracy in specialist referrals
✔️ 82% in diagnosis
⚠️ Triage still needs work

👉 Interview: www.helmholtz.de/en/newsroom/...

👉 Study: www.mdc-berlin.de/news/news/ai...

Evaluating large language model workflows in clinical decision support for triage and referral and diagnosis - npj Digital Medicine npj Digital Medicine - Evaluating large language model workflows in clinical decision support for triage and referral and diagnosis

full text: www.nature.com/articles/s41...

🩺 Clinicians mostly agree with what the models spit out—what mattered was whether its advice would guide safe care, not whether every word matched the textbook.

🤖 AI isn’t replacing clinicians; it’s fast becoming the teammate we’ve been waiting for. #HealthcareAI

🚑✨ Our new research in @springernature.com journal npj Digital Medicine shows that large language models accurately triaged, referred and even suggested diagnoses for 2,000 real emergency-department cases. 💊🏥

A microscopic image of cancer cells in shades of purple, overlaid with two hashtag boxes: #WorldCancerDay written in white letters on a dark magenta text box at the top, #UnitedByUnique in a light pink box at the bottom.

Each cancer is different, and so are the specific treatment options. "Onconaut," an #AI-based tool developed by #mdcBerlin's @al2na.bsky.social and his team, is designed to navigate through the maze of options more efficiently.

More: www.mdc-berlin.de/news/news/na... #WorldCancerDay #UnitedByUnique

Learn how to analyze spatial omics data! Join our computational genomics course in 2025. Application deadline 30th January. compgen.mdc-berlin.de
🖥️ 🧬 🧪

I'm happy to announce our next computational genomics course for 2025 🖥️ 🧬 🧪. The deadline for application is January 30th 2025. We will cover:
- AI-assisted data analysis
- Spatial omics analysis
- Multi-omics data analysis.
More info is here: bioinformatics.mdc-berlin.de/compgen/2025/

This is the last chance to apply for our computational genomics course, which includes an AI-assisted data analysis module. Deadline in 9 days. More info at compgen.mdc-berlin.de 🖥️ 🧬 🧪

What will you learn:
- Post-processing of spatial omics data
- Automated and manual image registration
- Niche clustering
- Multiple workflows for different spatial data types and data conversion
module by
@arturmansl.bsky.social

Learn how to analyze spatial omics data! Join our computational genomics course in 2025. Application deadline 30th January. compgen.mdc-berlin.de
🖥️ 🧬 🧪

GitHub - BIMSBbioinfo/flexynesis: A deep-learning based multi-modal data integration suite that aims to achieve synesis in a flexible manner A deep-learning based multi-modal data integration suite that aims to achieve synesis in a flexible manner - BIMSBbioinfo/flexynesis

3️⃣ Deep learning for genomic data Integration: Flexynesis simplifies cancer research by offering 1,000s of deep learning workflows, accessible to everyone🧠 github.com/BIMSBbioinfo...

2️⃣ AI-assisted data analysis with our multi-talent API 🚀💻 at rolv.io. This topic received multiple questions, as the possibilities for productivity are endless.

Onconaut: Navigating Personalized Cancer Therapies Explore personalized cancer therapies with Onconaut, a research tool for navigating treatment options.

1️⃣ Our AI deployments in healthcare: Multiple participants asked about safety and our future plans for deploying these technologies in clinics. Explore the potential with onconaut.ai and 2ndopin.io 🤖💡

Last week, I had the pleasure to give a keynote at the annual HIBIT Bioinformatics conference in Istanbul, here are the points that intrigued the audience most:

I'm happy to announce our next computational genomics course for 2025. The deadline for application is January 30th 2025. We will cover:
- AI-assisted data analysis
- Spatial omics analysis
- Multi-omics data analysis.
More info is here: bioinformatics.mdc-berlin.de/compgen/2025/

Join our upcoming workshop to learn more about our view on AI-assisted bioinformatics, multi-omics integration and spatial multi-omics. The course is for both R and Python users!!

Graphic with text: The title "ONCONAUT" is at the top, featuring an illustration of a space helmet with a DNA strand in the background [icon of the "onconaut.ai" program]. Below, the text reads 'ALTUNA AKALIN' with a small portrait of a person [Altuna Akalin] in the bottom left corner.

“Onconaut,” an #AI-based online tool that #mdcBerlin's @al2na.bsky.social has developed, helps oncologists and patients to find the most up-to-date and tailored treatment possible. @helmholtzai.bsky.social has now recognized it as the “Best digital resource.”

More: www.mdc-berlin.de/news/news/na...

I'm happy to announce our next computational genomics course for 2025 🖥️ 🧬 🧪. The deadline for application is January 30th 2025. We will cover:
- AI-assisted data analysis
- Spatial omics analysis
- Multi-omics data analysis.
More info is here: bioinformatics.mdc-berlin.de/compgen/2025/

I'm happy to announce our next computational genomics course for 2025. The deadline for application is January 30th 2025. We will cover:
- AI-assisted data analysis
- Spatial omics analysis
- Multi-omics data analysis.
More info is here: bioinformatics.mdc-berlin.de/compgen/2025/