Researcher at VUMC in the Ruderfer lab/Center for Digital Genomic Medicine. Predicting disease with EHR data and genetic information. Lots of ML. CS@mcgillu
@SashaGusevPosts Delivery and dosage testing is pretty tough in the early stages imo. The gap between some candidate drug working in cells/animal testing vs giving people that candidate drug in some form introduces a lot of interactions and barriers I'd guess.
I'm pretty confident that this "based in" feature is just where you were on the day they announced it, because mine says I'm based in Greece.
Coincidentally, I was on a trip in Greece during the week they announced that feature, which is interesting.
It seems like they're finally rolling out this feature lol, but I'm curious how the "based in" feature works.
This account was created in the US, and I am a citizen thereof, but most of my tweets have not physically been from that location.
@chrisalbon I've definitely had chatbots look like they solved a problem, only to find that they just told me that the feature I wanted to exist existed, when it did not in fact exist at all.
@colin_fraser This is exactly why precision is often a super misleading metric to focus on in medical spaces in particular when comparing results in papers. Very common for people to say that artificially balancing a dataset makes their metrics better, when it's really just changing a baseline
TBT has been having events for meeting up and chatting about biology, technology, and other fun stuff in several cities -- since I thought that sounded like a good time I decided to host one in Nashville! You can check out the event below if you also think that sounds like fun:
TechBio Transformers is coming to Nashville on Saturday, November 16th! Join us for a hike around Radnor Park, led by our regional lead, Ted Morley. Sign up here - https://t.co/yVM24a0j3D
@giffmana Incredible! I remember reading about this years ago, I feel like perhaps 2019, and thinking the idea sounded really interesting. I had completely forgotten about it!
Hugely deserved Nobels recently, I wasn't expecting such a focus on computer science. It's exciting to see the recognition of how computer science and specifically deep learning has been reshaping the scientific process over so many fields!
BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2024 #NobelPrize in Chemistry with one half to David Baker โfor computational protein designโ and the other half jointly to Demis Hassabis and John M. Jumper โfor protein structure prediction.โ
@littmath I think this is confusing intuitively for similar reasons to the Monty Hall problem. The fact that Bob has already observed a coin by the time that Alice does affects the optimal strategy for which coins to check, because it means the coin must be tails for her in certain subsets
@littmath Really interesting. You need to reach a certain level of flips such that there is a scenario where Alice is checking coins that Bob has already checked while neither of them have found heads yet, and as you increase the number of coins his advantage increases towards a limit.
@tcarpenter216 Seaborn is pretty solid in terms of python data viz, it's matplotlib under the hood but has some very nice looking visualizations with a lot of options.
ABSTRACT SELECT: @MorleyTheodore from @VUMChealth details knowledge-informed machine learning to identify undiagnosed patients with rare diseases in electronic health records at @AGBT Precision Health 2024. #AGBTPH
@dr_alphalyrae For the popularity, I think it's largely how many engineers are taught it in university before anything else, and engineering being very regulated with respect to tools compared to more general software engineering.