Ha, cool ill give it a try. Thanks!
24.04.2025 14:27 β π 0 π 0 π¬ 0 π 0@daninthenet.bsky.social
Senior Drug Designer | #AI Computational Medicinal Chemist | #medchem + #compchem | Also into #pottery and #oilpainting
Ha, cool ill give it a try. Thanks!
24.04.2025 14:27 β π 0 π 0 π¬ 0 π 0Anyone knows of any opensource/free tool for conformer generation faster than rdkit? And any more accurate? Or BOTH?!?!?! π₯² #compchem #chemsky #sbdd
24.04.2025 13:43 β π 2 π 1 π¬ 1 π 0Cool approach! More minds better results. www.nature.com/articles/s41...
23.04.2025 19:48 β π 1 π 1 π¬ 0 π 0Ahh, ai bot :)
19.03.2025 20:27 β π 0 π 0 π¬ 0 π 0Agreed! Iβve been a developer for many years. Itβs still fascinating how new thinking models are capable of producing really advanced code that require minor tweaks with good supervision. AND itβs only been a few months since first useful thinking models were released. what to expect in 1 year? π΅βπ«
19.03.2025 20:07 β π 0 π 0 π¬ 0 π 0Wow! 5x faster in 1 GPU and scalable to 32 GPUs is 160x speedup with existing hardware!
19.03.2025 14:17 β π 0 π 0 π¬ 0 π 02/2 Thereβs a super clever way to achieve much better performances! Just let me know if youβre interested in the topic and I can share more. #vibecoding #llm #cursorai
19.03.2025 11:19 β π 0 π 0 π¬ 1 π 0It turns out what I was trying to do with my side projects is called βvibe codingβ! Basically, I try to build full apps using LLM prompting and ZERO coding. Itβs rather impressive what can be achieved! Problem comes with large codebases due to limited LLM memory. But.. 1/2 #vibecoding #cursorai #llm
19.03.2025 11:19 β π 0 π 0 π¬ 2 π 0Big news: we are setting up a new non-profit organization to run bioRxiv and medRxiv. It's called openRxiv [no it's not a new preprint server; it's dedicated organization to oversee the servers] openrxiv.org 1/n
11.03.2025 13:20 β π 2571 π 849 π¬ 55 π 42Interesting new challenge similar in spirit to the well known CACHE!!
04.03.2025 12:59 β π 2 π 1 π¬ 1 π 0For all the gains AI affords in compound design and discovery, experimental validation still remains a bottleneck. But we're making that part easy for the cancer target TBXT with a suite of biophysical, in vitro and in vivo assay capabilitiesβplus prizes for potent binders: tbxtchallenge.org
22.02.2025 20:15 β π 2 π 3 π¬ 0 π 1Have you tried asking o3-mini to analyze a SAR table (from a screenshot) as a medchem?? I'm astonished! I was ZERO expecting this to work π± The results are REALLY impressive. #chatgpt #o3mini #medchem #chemsky #compchem #medchemsky
19.02.2025 15:46 β π 1 π 0 π¬ 1 π 0Sad news - Google goes to war: www.washingtonpost.com/technology/2...
13.02.2025 11:12 β π 2 π 0 π¬ 0 π 0I am using GPT4o for patent data activity extraction and it works very well! Much better than Excel OCR. #gpt4o #medchem #patentextraction
10.02.2025 22:17 β π 1 π 0 π¬ 0 π 0Iβm now playing with n8n to build AI agents. Itβs incredible how easy it is to set up an assistant. I managed to integrate it to Slack and also can chat with it through Telegram. I can ask it about my own notes using voice. π₯π₯ #ai #agents #n8n
07.02.2025 23:34 β π 2 π 0 π¬ 0 π 0And now o3-mini has also been releasedβ¦ and my first test is quite successful. I am prompting it to build a python interface for REINVENT4. So far o3-mini has performed best against deepseek and chat-gpt4. Best by far!
01.02.2025 16:12 β π 0 π 0 π¬ 0 π 0How do you imagine the new reasoning LLM models will impact drug design? Iβm starting to realize how important Agents will be in near future. They might become the new interface for everything we do. It is overwhelming to imagine all the possibilities.
01.02.2025 08:30 β π 2 π 0 π¬ 0 π 0Hi there! Iβm moving in. I need to rebuild the network I had on Twitter/X. I mainly use this to keep in touch with folks ans science in drug discovery in general. Structure based drug design in particular. Some support would be great! π
24.01.2025 07:04 β π 0 π 0 π¬ 0 π 0