#protein structure

[hellsite-proteins]

AlphaFold Nobel Prize!

Hey everyone :) this isn't a structure, but there is some protein news that is pretty relevant to this blog that I felt I had to share. This article gives a nice overview of AI-predicted protein structures and what sorts of things they can do for research. It's not too long, and I recommend taking a look

If you've been seeing my posts for any amount of time, I've absolutely given you a flawed view of how useful AF can be. Experimentally determining protein structures is a demanding and difficult process (I've never done it, but I've learned the overview of how x ray crystallography works, and I can only imagine how much work it would take). AI-generated structures are not going to make structural biology obsolete, but they are massively helpful in making predictions that go on to guide further research.

While in many fields (especially creative areas like art and writing) AI has significant ethical concerns, I feel like this sort of use of AI in science is an overwhelmingly positive thing. The data used to train it is publicly available, and science works by building on the work done by those before us. Furthermore, while AI may not be great at generating new ideas or copying humans, it is very good at sorting large amounts of data and using it to make predictions. It's more akin to very complicated statistics than an attempt at the Turing test, and in this case it is a valuable tool to expand the ways we can do science!

[hellsite-noobteins]

raw input: penis filtered input: PENIS it's just a wibble, not much else to say i think?

stick | cartoon | surface

colored by pLDDT

[cbirt]

OpenFold, an artificial intelligence (AI) research consortium, has announced the release of two new tools that aim to improve protein structure prediction: SoloSeq and OpenFold-Multimer. These open-source tools represent important advances in integrating protein language models with structure prediction software, as well as modeling protein complexes more accurately.

The SoloSeq model developed on Amazon Web Services (AWS) emerges as a groundbreaking protein LLM/structure prediction AI tool. It holds the distinction of being the inaugural fully open-source model in this category, offering critical training code. Now, organizations can leverage this code for fine-tuning or training new models with their proprietary data, marking a pivotal step towards collaborative advancements in the field.

The first tool, SoloSeq, combines OpenFold’s existing protein structure prediction capabilities with a new protein large language model (LLM). This integration eliminates the need for a separate pre-processing step to generate a multiple sequence alignment (MSA) that is required by OpenFold and other structure predictors like AlphaFold.

Continue Reading

[er-cryptid]

Intrinsic Membrane Proteins

Patreon

[bpod-bpod]

Building a Sperm

Combining protein structure predictions with cryoelectron tomography reveals proteins – novel and known – their structures, interactions and locations in intact mammalian sperm

Read the published research paper here

Image from work by Zhen Chen and colleagues

University of California San Francisco, San Francisco, CA, USA

Image originally published with a Creative Commons Attribution 4.0 International (CC BY 4.0)

Published in Cell, October 2023

You can also follow BPoD on Instagram, Twitter and Facebook

[hellsite-proteins]

this is it everyone. time to fully develop brain !!

fiveyearplanfckitweallfllydevelprainifstillinsanekillselfifnrmalstartpdcastwerernningtftimetfckitweall

letter sequence in this ask matching protein-coding amino acids:

fiveyearplanfckitweallfllydevelprainifstillinsanekillselfifnrmalstartpdcastwerernningtftimetfckitweall

protein guy analysis:

the confidence isn't great, but i do like how this one turned out. there are four nice helices, all arranged together in a way that looks like it could be a real domain. this is off topic, but sometimes i wonder how well the structures of protein interfaces that interact with other things are handled. for instance, if something was a transmembrane domain, or had a hydrophobic interaction with another subunit. i guess it would depend more than anything on how many similar structures there were, since i don't know how much the program looks at thermodynamics vs existing structures. i know at least that there are programs to predict transmembrane domains, i'm sure this is something i could research a little more instead of just making guesses. ultimately though, while there are some hydrophobic patches on the surface, they are weirdly dispersed in a way that makes me think this should be either a section of a full protein, or that this is actually just a mistake of language and computation that i have invested too much thought into.

predicted protein structure:

cartoon structure

surface, coloured by hydrophobicity (blue is most hydrophilic and yellow is most lipophilic, with white in the middle)

we're running out of time to fuck it we ball....

[hellsite-proteins]

Hi, im sorry if this violates no nsfw but i thought i might try coz its been an old meme; delete if you dont want to ans it, no pressure!

I've come to make an announcement: Shadow the Hedgehog's a bitch-ass motherfucker. He pissed on my fucking wife. That's right. He took his hedgehog fuckin' quilly dick out and he pissed on my FUCKING wife, and he said his dick was THIS BIG, and I said that's disgusting. So I'm making a callout post on my Twitter.com. Shadow the Hedgehog, you got a small dick. It's the size of this walnut except WAY smaller. And guess what? Here's what my dong looks like. That's right, baby. Tall points, no quills, no pillows, look at that, it looks like two balls and a bong. He fucked my wife, so guess what, I'm gonna fuck the earth. That's right, this is what you get! My SUPER LASER PISS! Except I'm not gonna piss on the earth. I'm gonna go higher. I'm pissing on the MOOOON! How do you like that, OBAMA? I PISSED ON THE MOON, YOU IDIOT! You have twenty-three hours before the piss DROPLETS hit the fucking earth, now get out of my fucking sight before I piss on you too!

tbh i'm not going to be super strict about the no nsfw thing, but i also don't entirely trust the internet and wanted to make sure no one would submit anything truly vile. this is more comedic than anything else though, so you're all good!

letter sequence in this ask matching protein-coding amino acids:

IvecmetmakeanannncementShadwtheHedgehgsaitchassmtherfckerHepissednmyfckingwifeThatsrightHetkhishedgehgfckinqillydicktandhepissednmyFCKINGwifeandhesaidhisdickwasTHISIGandIsaidthatsdisgstingSImmakingacalltpstnmyTwittercmShadwtheHedgehgygtasmalldickItsthesiefthiswalnteceptWAYsmallerAndgesswhatHereswhatmydnglkslikeThatsrightayTallpintsnqillsnpillwslkatthatitlksliketwallsandangHefckedmywifesgesswhatImgnnafcktheearthThatsrightthisiswhatygetMySPERLASERPISSEceptImntgnnapissntheearthImgnnaghigherImpissingntheMNHwdylikethatAMAIPISSEDNTHEMNYIDITYhavetwentythreehrsefrethepissDRPLETShitthefckingearthnwgettfmyfckingsightefreIpissnyt

protein guy analysis:

if you've been here long enough or have a general idea of what proteins are supposed to look like, i'm sure you could write this analysis from looking at the picture for about two seconds. all i have to say is this: why is it so spread out??? the answer, of course, is that all of the many polar and charged residues are way too hydrophilic to make anything real, and instead are just flopping around uselessly with all the water. proteins were never meant to have this much entropy, and this chaos is downright infuriating

predicted protein structure:

[hellsite-noobteins]

the missile is very tired. he is eepy. the missile has had a very long day of splashing bandits and wants to take just a small sleep. he eeby and neebies to sleebie. mibsile sleepy and need bed by time. the missile is currently experiencing critical levels of being a sleehjy little guy and needs to go to beb. he is retired and needs to slep. just a little sleejing time as a treat. mibsilelelele neebs to slek for twired boyo. just a lil guy. mibsipaleebeelee needs his beaty sleep. look at him go! he yawn bib cause he skeegy. neebs to falafel asleep. ni ni time. goodnight, mr the missile.

raw input: ↑ filtered input: themissileisverytiredheiseepythemissilehashadaverylngdayfsplashinganditsandwantsttakestasmallsleepheeeyandneeiestsleeiemisilesleepyandneededytimethemissileiscrrentlyeperiencingcriticallevelsfeingasleehylittlegyandneedstgteheisretiredandneedstslepstalittlesleeingtimeasatreatmisileleleleneestslekfrtwiredystalilgymisipaleeeeleeneedshiseatysleeplkathimgheyawnicaseheskeegyneestfalafelasleepninitimegdnightmrthemissile theres a shit ton of alpha helices and a few beta sheets, it looks mostly orange, kinda stringy around the edges, kind of standard for a peptomination. yet, i somehow feel like it's on the good side of standard? although from a certain angle theres like. a hole. why is there a hole did the missile do that

cartoon | surface cartoon hole | surface hole

colored by pLDDT

[cbirt]

Centuries after the first proteins were identified and characterized, their complexity proves a challenge for scientists all over the world. Their creation, evolution, and many variations are questions for which we have only just begun to uncover the answers. A new bioinformatics tool has now been developed that can provide researchers with a way to analyze data and map protein sequences and structures on a large scale. This has important implications for protein modeling and sequencing technologies.

Genomes provide the template on which all life is created and maintained. Despite the high accuracy of DNA replication mechanisms found in living organisms, certain mutations can still occur and be passed on. Such mutations can change the function and structure of the protein that the gene is supposed to encode. Nearly all Mendelian diseases are caused by variations in the proteome.

In lieu of the time-consuming and exhaustive process of determining protein structures and mutations in the lab, researchers have now turned to bioinformatics as an alternative method to quickly and cheaply generate highly accurate and precise protein structure models. Many protein modeling tools have been developed to predict mutations and their impact with as much accuracy and precision as possible. The majority of such tools use only protein sequencing information from both the protein as well as its variants in order to produce results. While such a method is advantageous in that it is applicable to nearly all genetic mutations, they don’t account for the structure of the mutated variants, resulting in a loss of valuable information that can be garnered from taking the proteins’ 3-dimensional structures into account. This is especially important as it allows for distinguishing between mere correlations in mutation occurrence and causative variations. Hence, the results obtained through such methods may not be as accurate or precise.

Continue Reading

[hellsite-proteins]

finally, i've nailed it on the timing of a post!

letter sequence in this ask matching protein-coding amino acids:

daystandIstillthinkImin

protein guy analysis:

it looks like AF3 just loves helices. in this case at least i do trust the helix, and honestly i'm pretty blown away by the accuracy of this, and the prediction is giving me >90% accuracy!! smaller structures do usually turn out better, but even then this feels to good to be random, so i did a quick BLAST search. sure enough, i found a 91% match for most of the sequence from a bacteria enzyme (specifically aldose 1-epimerase from Neobacillus drentensis), and another match for the last part. between these two proteins, we have nearly complete coverage of the sequence, and even if neither of these have solved structures, there are a number of other similar results, so this doesn't look too far off from evolved sequences (albeit just for a small peptide fragment). this lines up nicely with the strengths of AlphaFold, since it can use existing information to make a fairly trustworthy and reasonable guess. even then, this could only be trusted so much without any experimental structural information.

predicted protein structure:

cartoon representation, coloured by pLDDT

cartoon with side chains, coloured by pLDDT

highest % identity BLAST result

another strong BLAST result

pLDDT colour code

16 days to 2015 and I still think I’m in 2012

[hellsite-proteins]

Gay gay homosexual gay gay gay

happy pride everyone! i managed to get this one in before the end of june :)

letter sequence in this ask matching protein-coding amino acids:

Gaygayhmsealgaygaygay

protein guy analysis:

this turned out really well, with confidence ratings of 80% or more throughout most of the alpha helix. the side chains are also interesting to look at, because the glycine (G) and alanine (A) are tiny, while the tyrosine (T) is pretty bulky and aromatic. i don't really know what this little guy would do inside a cell, but its pretty fun to look at on the screen

predicted protein structure:

cartoon representation

cartoon with side chains shown as sticks

[hellsite-noobteins]

raw input: [the vaporeon copypasta]

filtered input: [the vaporeon copypasta but without the numbers, punctuation, spaces, and the letters b, x, j, z, o, & u. i'm not putting it here you can't make me]

a nice amount of secondary structure in the center, even including a few beta sheets! shame about the absolute spaghetti outside and through it. also 2 of the alpha helices look... crimped, in a word. a distressing peptomination for a distressing string of text. good thing neither are real like look how fuckign oragne it is. the surface view kinda looks like one of those things you get by pouring molten metal in an anthill and waiting for it to cool jesus fucking christ

(also. did you know ChimeraX has a Spin Movie feature? because i didn't)

cartoon | surface

colored by pLDDT

[hellsite-proteins]

can i request the lyrics to numa numa? english doesn’t seem to tend towards reasonable structures, at least that i’ve seen so far, so i’m curious if this is any more reasonable

sure thing! this song had a lot of u's and o's that i had to remove, but even after that it does look different from english to me with a lot of i's, m's and n's, although that could also just be the repetition. also this song is an absolute bop!

letter sequence in this ask matching protein-coding amino acids:

MaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaAlsaltsntenhaidcSitergiireameaprimestefericireaAlalsntePicassTiamdateepsisntvinicDarsastiinticernimicVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitaiVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitaiTesnsatispncesimtacmAliireameasntefericireaAlalsntiarasiePicassTiamdateepsisntvinicDarsastiinticernimicVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitaiVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitaiMaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaMaiahiimaiahMaiahmaiahaaVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitaiVreisaplecidarnmanmaieiNmanmaieinmanmanmaieiChipltasidragsteadinteiMiamintescdechiitai

protein guy analysis:

i absolutely love how this structure turned out. don't get me wrong, i don't think it looks like a "real, naturally evolved" protein in the slightest. there is a lot of diversity in how proteins look, but if this one were real, it would definitely be a pretty recognizable structure. this is mostly due to what i've decided to call "temu beta barrels", although i am taking alternate name suggestions. these rectangular tubes of beta sheets make sense from a computational perspective; there is a mix of polar and non polar residues so it makes sense to use beta sheets which have residues pointing in alternating directions, and parallel beta sheets are less stable, so more strands are needed, which works out well from the repetitive parts of the song. still, this looks nothing like beta barrels, which are one of my favourite structures and are made of a tube of twisted beta sheets. more so than most of the peptonominations i generate, i'm really curious about whether this would be stable in real life, and whether it would even be possible to synthesize and purify this.

predicted protein structure:

cartoon representation

cartoon with side chains

cartoon, coloured by hydrophobicity

examples of beta barrels

[hellsite-proteins]

my cells are more productive than i will ever be, and i commend the forces of nature for everything they have accomplished

letter sequence in this post matching protein-coding amino acids:

affirmatinsiamacmplerganismrtallyengineeredyncaringfrcesfnatreiamaprdctfillinsfyearsandtrillinsfdeathsiamildingamachinegreaterthanmyselfiamaletmakephnecallsandappintments

protein guy analysis:

there is a beta sheet and an alpha helix, but still a pretty standard amount of disordered loops around these structures. as much as i had complained about them before, i think its somewhat comforting to see some disordered loops along with the low confidence scores, since i don't really believe a lot of the structures that come out of these predictions. there are still some hydrophobic residues that are a little too exposed for my comfort, and like many of the peptonominations that this blog generates, i think this would be a challenge to try and express in cells. i don't think i've talked enough about aggregation, or how protein folding in vitro is different from in vivo, but i would be very interested to see how this interacts with a solvent or the mess of the cytosol

predicted protein structure:

cartoon representation

surface coloured by hydrophobicity

affirmations

i am a complex organism brutally engineered by uncaring forces of nature

i am a product of billions of years and trillions of deaths

i am building a machine greater than myself

i am able to make phone calls and appointments

[dr-afsaeed]

Would it be possible to kill bacteria without damaging their protein structure?

Would it be possible to kill bacteria without damaging their protein structure? There are several ways to kill bacteria without damaging protein structure or function. For example, consider the following frequent ones: 1. First, heat inactivation: Although heat may be used to destroy bacteria, it can also denature proteins if it is applied at too high a temperature for too long. However, some…

[hellsite-proteins]

i dont have a post but can you do this

heres the transcript

“Linda? What the fuck? Are you some fucking office secretary who just happens to be a middle aged white woman well into her 30s with a fucking demon of a child? Why in the ever loving shit did your parents name you Linda??? Like Linda is just Karen but somehow even more fucking stereotypical!”

peak writing

i'm not sure if this is a reference to anything or just thoughts on the name linda, but in my personal experience linda is just the boomer version of karen.

letter sequence in this ask matching protein-coding amino acids:

LindaWhatthefuckAreyousomefuckingofficesecretarywhousthappenstoeamiddleagedwhitewomanwellintoherswithafuckingdemonofachildWhyintheeverlovingshitdidyourparentsnameyouLindaLikeLindaisustKarenutsomehowevenmorefuckingstereotypical

protein guy analysis:

this was one of the rare structures that actually started to look worse as it was forming, and left me with a vile, unapproachable mess. i genuinely do not understand what rules the software had to break to even give me a pdb file, because there are side chains seemingly overlapping with each other and with the backbone. sure, no two atoms have their nuclei in the same space, but the bonds are so mangled that this could never stably exist. and this is even after the protein apparently severed some of the peptide bonds to make everything fit. apparently the rage against linda is strong enough to make electrons stop existing. this made me so mad i completely forgot that this travesty has some real alpha helices, if only to mock me. i am so upset that i had to look at this with my own eyes

predicted protein structure:

cartoon representation of the main chain

all side chains shown as sticks

close up of the center of the protein (so you can understand my pain)