ไม่สามารถเล่นวิดีโอนี้
ขออภัยในความไม่สะดวก
How to Use AlphaFold 3 for Biological Modeling
ฝัง
- เผยแพร่เมื่อ 16 ส.ค. 2024
- Link to AlphaFold 3 Server: golgi.sandbox....
Welcome to Bioinformatics Insights. This video demonstration will teach you the usage of AlphaFold3 developed by Google's DeepMinds collaborated by Isomorphic lab. You will also learn how to analyse model quality generated by AlphaFold3.
#google #deepmind #autodockvina #alphafold3 #alphafoldserver #modeling #bioinformatics #bioinformaticsforbeginners #bioinformaticstools #biology #biologyinsights #neuralnetworks #deeplearning #ai #protein #ligands #ligand #ions #oligomerization #molecular_biology #molecularbiology #proteinresearch
![[TALK 21] AlphaFold: Use and Applications - Sami Chaaban](/img/n.gif)
Hey, I found a mistake in the video. When you add the DNA, you add 2 copies of it. But DNA is anti-parallel. So you would need to add the reverse complement (not the second copy) in order to simulate the double helix. I am not sure why you got the double helix in the end (palindromic sequence maybe?) but there should be the reverse complement.
@nataliyateteneva1679 you are right to the point the AF reads 5' to 3' but AF is enough smart, if you ask for two copies, it will place the second one on 3' to 5'. You can model a DNA and then check orientation of both the chains in any visualizer. Thanks for asking
If I have multiple proteins sequence and a medication sequence how I can add this to alpha fold or should I use decking program? After downloading the 3d shapes from alphafold
I didn't understand the term medication sequence. Please elaborate, thus I may provide a reasonable reply
Sir very informative. Regarding to interaction in the case of dimer or trimer or hetero macromolecules. Can we dock them or just to model here via alphafold3
Amir, AF3 handles oligomeric modeling better than other available packages.
❤ Thank you so much😊
You're welcome 😊
I want to do protein-protein interactions (molecular docking). I can use AF3 for protein modelling. Can I use AF3 for molecular docking? If yes then how? If AF3 can docked both receptor and ligand protein then in which tool should I visualize docked complex? Previously I docked in ClusPro and then visualize complex in ChimeraX but chimeraX considered docked complex as a single protein. How PyMol showed both docked proteins as an individual but I don't know how to find interacting residues. Jazak Allah
Dear, all the questions you dropped here are already addressed on the channel. Let me answer you again here.
1. Yes you can use for AF3 also AF2 for protein protein docking, and the structure accuracy can be accessed from pTM in case of more than 1 protein.
2. How, dear I have covered the video already, where I have explained all the metrices etc. th-cam.com/video/iGGby6zavAY/w-d-xo.htmlsi=WE9R1-O8HFagUtaS
3. AF3 can not dock all the ligands, only few. You will learn in the above.
4. Docked complex as a single protein.. I think you didn't observe, normally you will see different chain of docked proteins in a complex... But if not, you can use chimera editing option and assign them different chain numbers.
5. Too much are covered on PyMOL on this channel. But if you want to see the interface of the complex in PyMOL, please consider the following tutorial.
th-cam.com/video/6hEDqNRDdJo/w-d-xo.htmlsi=BSxq0COw7_-uvoEf
Hope I have answered all your questions.
Thank you 🙏🏻
My pleasure.
good
Thanks for watching
How can I add an internal interaction of my protein to AlphaFold prediction?
Gavin please elaborate your query. What exactly you want to achieve from AlphaFold? Also what you mean from internal interactions. There are certain things beyong the scope of AF. Upon your query, I will be able to respond you.
@@Bioinformaticsinsights Thank you for your answer. Basically, I have a protein (genetically) connected with a peptide on the N-terminus and with a nanobody on the C-terminus. I want to let these 2 outer domains make an internal interaction with each other to see what their effect is on my internal protein. I hope this makes it more clear.
Got it. Now you have two options the one suits you can be tested.
1. Just copy your peptide sequence and paste it in the start of the protein. In this case AF may take it as a part of the protein. If it is reported that this peptide interacts the N-terminus of your protein then follow option 2 below.
2. Include the sequence of your protein and Add another line, as I did in video for DNA etc. but you have to select it protein too. Now paste the sequence of your peptide. I believe AF is enough smart to keep your peptide on the N-terminus.
Please let me know if you have further questions.
Cheers
Really great + thanks
Glad you liked it!
Hi, very nice tutorial. If I want to model the interaction between a DNA and a mutant protein (vs WT), how can I change the amino acid sequence? Thanks in advance!
@huiruchen7808 for the wild type you need to copy and paste as mentioned in the video. While modeling the mutant interaction, first paste the wild type sequence, now click inside the protein tab and change the specified amino acid. Also, you can do this way, paste the wild type sequence into wordpad or notepad. Change the sequence there, and paste it to the protein tab. Now confirm it before submitting job whether a correct amino acid has been changed or not.
Goodluck
@@Bioinformaticsinsights Hi, Thank you so much! I have several questions. You mentioned that if the score of pTM is below 0.5, you need to the step, turn on the seed, and increase the seed to get another modeling result to see whether you can get a higher score of pTM. Could you please tell me how to do that? When I export the results, there are 5 predictions, how do you usually interpret the 5 predicted structures or what kind of value is the most critical parameter to decide the most accurate structure? Thank you again!!! :)
Answer 1: at the end i have mentioned from where you can manipulate seed information.
Answer2: All the metrices to take decision are explained in the video.
Hi, Thanks for the video, but When I am downloading the job files, I am getting results /output in Jason. How can I access the pdb files? any idea? many thanks
@lakshyarajkhatri Dear, you will receive a .zip file upon download. Once you unzip it, you should see .JSON and mmCIF files. The mmCIF files can be viewed in pymol and others visualizers too. Once it it open, you can save it in PDB formate.
Many thanks for the info
How I can download the predicted structure and use in pymol
Please watch the complete video, at the end I mentioned the place from where you can download.
How can I used sugars as a ligand as it is not in the list of ligands?
As I have mentioned in the video, AF is not a docking program. To dock sugar, you have to consider docking programs. I have already explained tutorial of AutoDock Vina by two different methods. You can use it.
Sir will you continue to upload CADD more videos?
Yes, that is not complete yet. To upload some refreshing and new stuffs, I placed this video.
@@Bioinformaticsinsights Thank you so much 😊