Organic Chem Γ Material Science x Entrepreneur x Innovator | PhD Chemistry and Materials Science (NYU)π§ͺ Biotech and wireless-tech https://teapaw.org/t.co/YTp8j0TJzT
New York, USA
Joined July 2014
- Tweets1.6K
- Following1K
- Followers295
- Likes3.2K
Thanuka Dananjaya π retweeted
AI promises to cure disease, but can frontier models interpret the images biotech/pharma scientists handle every day?
To evaluate this, we built the ATLAS Visual Life Sciences (VIALS) benchmark: 161 visual interpretation tasks from professional life sciences workflows including gel blots, plasmid maps, molecular structures, and phylogenies. These are the images/results that experiments actually produce, not polished figures from papers or textbooks.
We benchmarked todayβs 10 leading multimodal models. Opus 5, GPT 5.6, and Gemini 3.7 all fail to crack 27% accuracy, and the other models like Kimi K3 & Grok 4.6 perform much worse.
In contrast, PhD-level biomedical scientists find these visual interpretation tasks straightforward. AI that cant interpret such images will have limited utility in life sciences R&D, where such artifacts are central to how scientists reason, communicate, and make research decisions.
We hope VIALS drives progress toward models that can.
AI can cure cancer... by figuring out how to generate revenue to fund scientists working. The problem has always been funding and not the lack of creativity in experimentation.
Thanuka Dananjaya π retweeted
Had a great time recording this with @aplooby28 - more than 90 min of chilies, chemistry, careers and more!
Thank you to Synthesis Workshop @MatthewHorwitz1 for supporting this great podcast! πΆοΈπ§ͺ
Join us today on The SW Pod with @aplooby28 as we sit down to chat with the legendary Scott @BagPhos Bagley!!!
Spotify: open.spotify.com/episode/7xSβ¦
YouTube: youtu.be/Xmqpk0wFPTg
Thanuka Dananjaya π retweeted
Live on @RowanSci: new molecule builder updates that make it easy to move and rotate structures. This makes it much easier to set up a TS opt of a reaction, optimize systems of molecules in a given arrangement, etc.
Try it out at labs.rowansci.com on any submission page.
π§΅New paper alert!
Published in @BioconjChem this week, here we present chemistries and strategies to produce and isolate novel hydrophobic-DNA conjugates.
#DNA #synthesis #chemistry #Openaccess #NYU #nanotechnology
tinyurl.com/y8nap2ff
(1/10)
This is the first paper in a series that would be coming out within the next several months, and I am happy to have worked on projects where we really tested the limits of solid phase chemistry, hydrophobic/philic compatibility, and detection of highly unnatural conjugates
(9/10)
Should out to Prof. James Canary and Dr. Ruojie Sha for all the guidance and providing an amazing support system. Big congrats to Chufan Yang for spearheading Stille coupling-based approach and this paper. @USNavyResearch for the funding support.
(10/10)
Thanuka Dananjaya π retweeted
Excited to share my latest research in @ajerschow lab on the conformational dynamics of NADH in different solvent environments and the impact of metal ions! Check out the further details in my presentation β¬οΈ BTW hope you like the background music I chose π #GlobalnmrOC2024 #nmr
Thanuka Dananjaya π retweeted
Out today: ADCs in combination with high-throughput experimentation, microfluidics, and flow, demonstrating the usefulness of non-toxic linker-payload model compounds for analytical, process, and technology development via #merckchemistry & #biologics collaboration. Enjoy!
Online today in #OPRD! ADC conjugation empowered by data-rich technologies, continuous flow and high-throughput experimentation, using tool molecule collections of non-toxic linker/payload mimics. #merckchemistry @emmertlab @cecibottecchia Read it here β€΅οΈ go.acs.org/abv
Thanuka Dananjaya π retweeted
I benefited so much from others helping me this way when I was early in my career
Iβm so impressed with the level of science in the grad student/post docs/ug I see at ACS events
at that age I was in my βcan I extract with MeOH?β phase
Last week @BagPhos walked up to my poster, saw a reaction I was struggling to get useful yields on for over a month, spent all of maybe 5 seconds thinking about it, rattled off a list of reagents, sent me a citation and then confidently said "that'll work"....