AlphaFold3 is out!
This a diffusion model pipeline that goes beyond what AlphaFold2 did: predicting the structures of protein-molecule complexes containing DNA, RNA, ions, etc.
Blog post: https://t.co/0HEgbPwUOC
Paper: https://t.co/00Lbw5xhPQ
A quick thread about the method↓
@Carnage4Life *he says, as Nvidia continues to hire record numbers of developers*
Imagine if we stopped learning mathematics when calculators were invented 😳
@ATinyGreenCell Instead, we learn how to analyze these large pieces of raw data.
This is the same philosophy that people attempt to train AIs with. Who cares if it knows the entire sequence of the human genome when it can tell you exactly how to get the information you want from that genome?
@ATinyGreenCell A list of boiling points and the periodic table are things that people do memorize and information that this model would be trained on, because they have direct uses in various fields. There’s no point in memorizing sequences, because they have no direct use.
@AnthonyLeeZhang Lol I don’t share the same opinion, but you can totally use either R or Python for all data science tasks with *nearly* identical outcomes. The best data scientists I know tend to use R for data wrangling/vis and Python for production, but that’s certainly not necessary.
Enough bioinformatics skills to build an app! After writing most of the week, I can't wait to publish this tool as a helpful resource to the #microbial#CRISPR community -- especially for our inducible B. burgdorferi CRISPRi system at https://t.co/ipbat5rujQ
@WhatIsAWomann@WordWanderer@JeffJacksonNC Interesting that you don’t think the strictly partisan decision which aligns with your opinions (that broke a bipartisan agreement) is a slippery slope... sounds almost like you’ve already slipped
@WhatIsAWomann@WordWanderer@JeffJacksonNC It’s almost like those Supreme Court Justices took their sworn oaths of “justice without respect to persons” seriously… and you’re the one who is, in fact, making a false comparison 🤔