Sheffield-iGEM / .github

0 stars 0 forks source link

Check AI guide from Felix against the paper! - Rough Draft #70

Closed KOXLEY2 closed 2 months ago

KOXLEY2 commented 3 months ago

https://pubs.acs.org/doi/10.1021/jacs.3c10941 ^ See what they do here to improve things

https://www.science.org/doi/10.1126/science.add2187

ikesadbot commented 3 months ago

I've added a TL:DR to this paper in the Zotero, under the "Synthetic tools for protein engineering". Seems as though the inpainting and PMPNN had huge results for redesign of myoglobin, but they didn't use it for the redesign of TEV and managed to get a crazy redesign that was twice as catalytically efficient and more than double the melting temperature!

TheLostLambda commented 2 months ago

Is this finished? Can I get a look at it? I'd be happy to typeset it in something like https://typst.app/ if we want a nice PDF? Might be worth also adding some of the inpainting stuff if we've not already, and perhaps adding in the ConSurf stuff, then we can advertise a Webinar for the other iGEM teams!

Gheist404 commented 2 months ago

@TheLostLambda yes it has been written up but should probably be checked through we did a mini test on how useful it is today here's a link to it https://docs.google.com/document/d/1s5mNQ9qOhT8wngCvj5V0P9svfnyygt7iI7JmMDGnHRM/edit?usp=drive_link it can also be found in training material/protein bioinformatics & design/protein redesign and it's called AI redesign cycle and that sounds great. Also the aim for this is just downloading and going through the first loop I am going to make a separate document going through actual details of the cycle but will get a final clarification from Felix before doing that!!!

Gracemarsden04 commented 2 months ago

@TheLostLambda Look at this!

TheLostLambda commented 2 months ago

Generally looks really good! Four things we should probably do to get it ready to publish and share with others: 1) Add in some sections about AlphaFolding with ions, and a section about picking out conserved residues using ConSurf to know what to fix at the end — also anything you did in ChimeraX to get those functional residues! 2) Add some more explanation in some sections, making it clear why we're doing each step 3) Add an introduction / background section for talking about why people might want to use these tools, what they can and can't do, and perhaps a quick overview of how they work (very broad strokes) 4) A final bit of editing — making sure the language is clear and concise for a wide audience of people with English potentially as a second language!