#Generative Model
Explore tagged Tumblr posts
cbirt · 6 months ago
Link
A part of systems biology is ‘designing RNA structures,’ which entails planning and organizing molecules of RNA into custom 3D formations with particular uses. RNA plays a dual role in cells, encoding genetic information and performing structural functions. Proteins are somewhat easier to study and model because of the precision nature of this macromolecule, while RNA conformations pose a special challenge because of their flexibility and multiple interactions. Deep learning and generative models have added a new generation to RNA structure design that bids new horizons for biotechnology and medicine. Enter RNA-FRAMEFLOW, a novel way for de novo 3D RNA backbone design that leverages flow-matching principles to create accurate and functional RNA structures.
The problem of designing the backbone of RNA is rooted in the fact that RNA undergoes conformational changes and because RNA interactions are complex, including base pairing and stacking. Proteins can be exemplified as relatively simple molecules by comparing them to RNA structures in which each nucleotide contains 13 atoms that require precise parameterization. This complexity, along with the loss of explicit physical interaction representations in current deep learning models, makes RNA structure prediction not only less precise than protein structure prediction but also nuanced. Solving these issues is essential for evolving the RNA modeling and design.
Continue Reading
1 note · View note
sevilai · 20 days ago
Text
Tumblr media Tumblr media Tumblr media Tumblr media
You ever have an idea so dumb you drop everything to draw it
10K notes · View notes
mindsynthcreations · 2 months ago
Text
Tumblr media
9K notes · View notes
humming-fly · 2 months ago
Text
Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media
He was feeling left out
and the higher rez stills, since gifs always export as if you're sending messages through a metal can~
Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media
4K notes · View notes
nmdry · 4 months ago
Text
Tumblr media
3K notes · View notes
hymyarts · 7 months ago
Text
Tumblr media
"We Myrmidons are not the sentimental type."
8K notes · View notes
cozylittleartblog · 6 months ago
Text
Tumblr media Tumblr media Tumblr media Tumblr media Tumblr media
Columbo and the Knight (1984)
put me in the universe where Columbo ran through the 1980s and had a crossover episode with Knight Rider. I think they deserved it, and I am not just saying that because they're my two favorite Old Shows. @telebeast wrote a little fanfic blurb about it and I HAD to visualize it into a comic (which is also the longest comic I have finished thus far at five pages...), so writing credit goes to them.
Autism W!
#columbo#knight rider#art#michael knight#kitt#comic#highlight reel#crossover#telebeast#there are two small easter eggs here. can you find them. they were somehow not Entirely lost when i resized these for the public#this is what i mean when i say I Draw And It's Everyone Else's Problem. look at my INCREDIBLY niche crossover comic boy#if the knight rider fandom has like 12 people in it. how many of y'all have seen columbo#this comic is for like 4 people and me and phoenix are already two of them#niche is my specialty lets be real. weird niche obscure shit and ships nobody's paid attention to yet#not to suggest this is ship art. columbo has his wife and michael has his car lmfao#stylizing real people is EXTREMELY hard btw sorry for when they get off model. its partly a 'better imperfect than never finished' situatio#cant tell you how much i redrew some of these panels. weeps#this took me 2 weeks but i think i thumbnailed it all in may and the ideas been rollin around in my head since march#is anybody good at editing. please edit michael and columbo into an image together like its a screenshot. NOT generated. edited.#it would be so cool#ive drawn columbo a lot but i haven't drawn a lot of michaels. i was learning things about his outfit AS I WAS DOING THE DAMN#COLORS ON THIS. all the lines done. it was too late to change anything. i did all the lines and colored page by page#i realized my mistakes on like page 3. 1 and 2 were already done. it was Too Late.#imagine it though. them working a case together. switching between the more serious tone of columbo vs the goofier#action antics of michael and kitt. columbo being so impressed by Modern Technology. there's more i could say but phoenix may write#more of this crossover and i don't want to spoil it :'3#there's opportunity here though i swear. there's gold to be dug.#i like how kitt gets shading but columbo's junker peugeot doesn't. kitt looked wrong without any. columbo's car is matte and dirty#i also applied effects to this to make it look a little film-grainy and VHS like. some CRT TV vibes#the only question left is. did they put knight rider into columbo; or columbo into knight rider 🤔
3K notes · View notes
ozzieort · 2 months ago
Photo
Tumblr media
Hey! I'm ozzieort. I create art in a cartoonish style. Check out my page and have fun :) If you like my art send me a message! I do custom commissions Support me on my Patreon
2K notes · View notes
lanalee193 · 1 month ago
Text
Tumblr media
1K notes · View notes
cbirt · 11 months ago
Link
Dynamic structural ensembles play a major role in the biological functions of proteins. For the purpose of learning and sampling the conformational landscapes of proteins, researchers create a flow-based generative modeling technique in this work. Sequence-conditioned generative models of protein structure are produced by optimizing the highly accurate single-state predictor AlphaFold technique within a unique flow matching framework. This approach outperforms AlphaFold with multiple sequence alignment (MSA) subsampling in terms of precision and diversity when trained on the PDB. For proteins that are not yet known, it precisely records positional distributions, structural flexibility, and higher-order ensemble observables. The technique has the potential to serve as a stand-in for costly physics-based simulations because it can diversify static PDB structures with faster wall-clock convergence.
In order to perform their biological tasks, proteins frequently take on complex three-dimensional structures as components of structural ensembles with unique states, collective movements, and chaotic fluctuations. The intensity and selectivity of molecular interactions, including those involving transporters, channels, and enzymes, are controlled by these structures. While deep learning techniques such as AlphaFold are excellent for single-state modeling, they are not suitable for taking conformational variation into account. Structural biologists might benefit from an approach that reveals underlying structural ensembles and builds upon single-structure predictors.
Continue Reading
2 notes · View notes
xanna-tose · 2 months ago
Text
Corseted Fashion 💖💝💖
2K notes · View notes
mindsynthcreations · 3 months ago
Text
Tumblr media
13K notes · View notes
masha-k-ia · 4 months ago
Text
Tumblr media Tumblr media Tumblr media
Do you like Emo?
Visit Patreon exclusive gallery
2K notes · View notes
naughty-salsa · 5 months ago
Text
Tumblr media
3K notes · View notes
nmdry · 4 months ago
Text
Tumblr media
3K notes · View notes
ozzieort · 1 month ago
Photo
Tumblr media
Hey! I'm ozzieort. I create art in a cartoonish style. Check out my page and have fun :) If you like my art send me a message! I do custom commissions Support me on my Patreon
1K notes · View notes