Mike A. Nalls
@mike_nalls
Followers
1K
Following
4K
Media
235
Statuses
1K
Founder at @DataTecnica, team lead at NIH’s CARD. Data science | healthcare | biotech. “oh, you really still skateboard?” Views/tweets my own.
Washington, DC
Joined August 2018
https://t.co/nPhr0hPDlA Check out our progress on cataloging all the #openscience and #FAIR resources in Alzheimer’s and related dementias! #knowledgegraphs #AI #dementia. Give it a test drive and let us know what we can add! Massive work from CARD and @DataTecnica teams.
datatecnica.com
Introducing the CARD Catalogue Beta: Advanced Federated Resource Browser for Alzheimer's Disease Research We are pleased to announce the beta release of the CARD Catalogue, a comprehensive federated...
0
0
2
Hot take maybe but J. Fonseca is gonna be stacking up trophies and leading the new “Fab Four” in a year or two, if not sooner. Tennis’ new Senna.
0
0
1
Can LLMs perform reliably as biomedical data analysts? TL;DR: We created the first benchmark designed to challenge LLMs ability to apply scientific reasoning in text-to-SQL generation over biomedical databases, revealing a 30-40% gap between SOTA models and expert performance
1
1
1
📄Read the preprint: https://t.co/hl9BPIxT5a 📊Dataset: https://t.co/y6MPciJSwb 💻Code: https://t.co/Qlbl33E0RL Thanks to my teammates at NIH/CARD and @DataTecnica including Maya Willey, Adi Asija, @owenbianchi, Chelsea Alvarado, @mike_nalls, @DanielKhashabi, and @FarazFaghri
0
1
1
Long-form inputs (e.g., needle-in-haystack setups) are the crucial aspect of high-impact LLM applications. While previous studies have flagged issues like positional bias and distracting documents, they've missed a crucial element: the size of the gold/relevant context. In our
3
21
52
Michael Schumacher has signed a helmet for Jackie Stewart's dementia charity with the help of his wife Corinna ❤️✍️ The helmet will be auctioned off after Jackie's tribute lap in Bahrain today.
87
1K
22K
Yuki for Scuderia Torro Rosso?
0
0
0
Summary blog here if you just want the highlights ... https://t.co/Y99tMfnD43 ...
datatecnica.com
Proud to announce the preprint of next generation of Parkinson’s disease genome-wide association study meta-analyes is here . This study really shows the power of unified data generation and analysis...
1
1
6
Big Parkinson's news. Big open science. Check the preprint for the new GWAS from 100s of collaborators across the globe led by Hampton from our team @DataTecnica. >1.8 samples and 157 risk variants.
3
28
167
“It’s the water” … 2026 is our year.
0
0
0
Anyone know anyone in sports analytics (particularly Motorsport, boxing or tennis)? Would really like to buy them a coffee and learn about their realtime data analysis set ups and automation.
2
2
3
Forza 🦊 🥕 !
Jannik Sinner. Holger Rune. Point of the tournament. Unbelievable stuff
0
0
1
Huge collaborative effort from the NIH/CARD and @DataTecnica teams, in particular @owenbianchi , Maya Willey, @mkoretsky1 , Chelsea Alvarado and @FarazFaghri .
1
0
0
My favorite test for this was asking it "which SNPs and alleles make the compound genotype for APOE4?" (I always forget this and have to spend like 10 min trying to find my notes or an old paper on this) ... https://t.co/qi3azhP0Mr nailed it in < 30 seconds.
1
0
0
CARD-AI automates a lot of the tedious work a junior analyst would do in terms of queries like looking up and synthesizing data across many different tables / databases.
1
0
0
https://t.co/LQvgr4o3MM 👀 We made an agentic AI research assistant for neurodegeneration and are rolling it out at NIH's CARD. The attached pre-print is the benchmark for it and it's pretty epic in terms of performance compared to off the shelf LLMs w/o custom curated data.
1
0
0