PyRosetta Tutorial

[@bioflows1321]

Hello Sabri. Thank you for this great jumpstart to Pyrosetta. I don't think that Glycine residue 500th is 66th just because, Pyrosetta starts from zero or one. because if it starts from zero, Gly should be 499th not 66. Pyrosetta gave 66 to Glycine which is 500 just because any Residue internally in pyrosetta is structured as such it is identified by its index as well as its chain number. For instance, the "pose" variable in python is an internal data structure which maps to the cleaned 1YY8 pdb file in which it is composed of 4 chains , A=213, B=221, C=213 and D=221. now let's do the math, your Glycine residue in the sequence at 500th location corresponds to chain C in your cleaned pdb file. now A+B chains is 213+221= 434 . Now your glycine should be at 66th location in Chain C because 500-434 = 66. So Pyrosetta identifies every residue not only by its index but also by its chain. so the physical location is a tuple composed of its index and the chain where it is resides. I hope this helps.

[@معجبالقحطاني-ب7ج]

يعطيك العافيه ي دكتور انت افضل استاذ درس مادة الاحياء في العالم

[@nataliyagolovach3063]

Thank you, the tutorial is very helpful for PyRosetta initial introduction!

[@melevarck]

Thank you very much for this tutorial! You effort made it much easier for me to get started with PyRosetta. Are you maybe taking suggestions or request on new tutorials? I would really like to see a tutorial about RECON multistate design using PyRosetta. I have been struggling to try to develop a script based on the official documentation and the original paper's xml script but I can't get it to work at all... I have no idea on how to pass multiple poses (I've tried using a vector but it failed) to the Movers's.apply() nor how to take this output and pass it to the subsequent Movers.

[@cpav9062]

Thank you for this tutorial, I really enjoied it! Do you know if the whole project is already ported from C++ to python? I mean, are the NMR-based protocols implemented in pyrosetta as well? BTW, I think that at 47:53 weighted_score = weight * raw_score

[@chenlinlu6990]

Very nice tutorial. Thanks

[@alon800]

Great Tutorial. Thank you very much!

[@chriszhang7133]

Thank you very much for sharing

[@yecomlekci5589]

Thanks a lot for this tutorial. Is it possible for you to make a tutorial for protein-ligand docking by using Pyrosetta?

[@SariSabban]

Yecomlekci I will. When I find the time. 🙂👍 hopefully soon 🙂

[@yecomlekci5589]

@@SariSabban thank you so so much again waiting for it with a hope :) I can do it with vina but want to confirm my results with pyrosetta from the window of whether it can dock or not :) I can prepare my ligand as params file and also my protein with chimera but docking part is kind a hard :D

[@patchaisupa8141]

Thank you. Great tutorials! It's really helpful

[@soumenbarman4661]

You are great sir. Sir is it possible to make a video on rotamer library creation with a noncanonical amino acid in rosetta or pyrosetta.

[@Xahra92]

Do you think Pyrosetta can be used to perform all functions of rosetta? I just joined a PhD program where I will be using molecular modeling programs. My professor said that it is better for me to learn rosetta instead of Pyrosetta. What would you recommend though? P.S. Thank you for posting this video!

[@SariSabban]

Hi, sorry for the late reply, I did not receive a notification of your post. Rosetta is the original software where everything you need is found, plus you can use Rosetta script to invent new protocols. PyRosetta lags a bit behind Rosetta, so you may not find all the protocols available, or you might find some protocols that work in Rosetta but do not work fully in PyRosetta. Also you can compile and use Rosetta on a supercomputer while PyRosetta still has difficulty correcly compiling on an old architecture usually found in supercomputers. I would take your supervisor's advice and learn rosetta, and afterwards if you have time learn pyrosetta (it is much easier and you can learn it in a week once you understand how rosetta works)

[@judochemistry]

Thank you very much for your great tutorials about Rosetta. I am a Mac user and am not sure if I should start with Rosetta 3.12 or PyRosetta. Do you think PyRosetta will be easier for me to start with?

[@SariSabban]

Both should work fine in a mac. Other than installation the commands should be the same.

[@joshtseng6001]

I have tried to do exactly what you did, but when I got to the point where I tried to extract the secondary structure, and it said "NameError: name 'get_secstruct' is not defined. Do you have any suggestion on what might be going on? I've checked the spelling and it was all correct.

[@SariSabban]

Josh Tseng This is unfortunately one of the challenges of using PyRosetta (finding the path to the function) Try this: from pyrosetta import * from pyrosetta.toolbox import * init() pose = pose_from_pdb('filename.pdb') ss = pyrosetta.rosetta.protocols.membrane.get_secstruct(pose)

[@joshtseng6001]

Sari Sabban Thanks for replying, I was able to figure it out by using pose.disply_secstruct() , to extract the secstruct. Thank you so much

[@kamilkaya5367]

How can I use Rosetta's energy function. I have got a project about Protein Folding problem using Meta Heuristic methods but I couldn't figure out Energy Functions?

[@offmeds2nite]

Very useful thanks

[@jovanastelkic9754]

Thank you for this great tutorial! Can you pleas help me to get secstruc , when i run the same as in tutor , it comes out an error" name 'get_secstruct' is not defined" only way to successfully get secstruct is ss=pyrosetta.rosetta.protocols.membrane.get_secstruct() or DSSP = pyrosetta.rosetta.protocols.moves.DsspMover(),but neither of them gives amino acid list in the same output with predicted secondary structure, output contains just secondary structure. also I would appreciate if you can tell me how did you set up terminal so that it has text editor on one side and executor on the other. I am on ubuntu 20.04., also have pyrosetta4, and vim but have very poor experience in both. thank you

[@SariSabban]

Regarding the terminal, it is easy, I use tmux. It allows you to tile the terminal. I recommend vim as an editor because it is powerful once you get use to it. Regarding the secondary structure: are you trying to print the secondary structure of a protein from a PDB file? Or predict the secondary structure from a FASTA sequence? These are two different things.

[@SariSabban]

Also send me the full code to see it. Sometimes the error is somewhere else.

[@jovanastelkic9754]

@@SariSabban thank you for your reply! i tried from pdb get_secstruct(pose) , but it said that get_secstruct is not defined, later on i found the other way to get it but it prints just H,L...so it is hard to track which amino acid correspond to which part of the structure.

[@SariSabban]

@@jovanastelkic9754 yes it prints it as a string. You will have to parse this string using python string manipulation and organise the secondary structure sequence with the amino acid sequence

[@jovanastelkic9754]

@@SariSabban thank you a lot! Best regards!

[@kamilkaya5367]

Hello. I want to use rosetta's energy function in order to solve protein folding problem using meta heuristic optimization simulated annealing. When make changes with dihedral angles(phi, psi, omega) how can I prevent clashing? Can Rosetta handle those situations? Also When I use score function how can I get the Energy value for specific type of sequence with given dihedral angles?

[@SariSabban]

This might be difficult to answer in a comment. But you can pass in a structure and print out the energy score. I show that in the video. Then it is up to you how to manipulate the structure and re-print the score to see how it changes. There is a PyRosetta folding with simulated annealing tutorial from the gray lab website. Go to PyRosetta.org and look for the folding tutorial and the final script code. I hope this helps.

[@TrishitBanerjee]

Hello Professor! I have been wanting to use Foldtree but I don’t use PyRosetta. I am using the general Rosetta. For foldtree, I understand that I would need to number my residues according to Rosetta numbering and not PDB numbering. Is there a way to do so? Thank you!

[@SariSabban]

Trishit Banerjee to be honest. I haven’t used fold tree before so I can’t help you with this particular question. Sorry.

[@TrishitBanerjee]

Thank you so much professor! :)

[@alioz63]

الله يوفقك ❤️

[@ruinninja90]

How do you see the residue you mutated in pymol? Can you save the protein on pyrosetta as a pdb to view any changes?

[@SariSabban]

You can mutate a residue in pymol. But it is better if you mutate it in PyRosetta because afterwards you can perform a relax step to get rid of any clashes, Search google for PyRosetta demo gray lab tutorial. Look at mutate-residue() You can always export a pose by pose.dump_pdb()

[@ruinninja90]

@@SariSabban Thank you! Also is there a way you can insert/delete an amino acid into a residue sequence? I have tried looking everywhere and I have no idea how to do it

[@nasimsoltani3768]

Hello. Assume I have two proteins that have known tertiary structures. Is there a way that I can load these two proteins in pyrosetta, make some restraints between residues of these two proteins, and then use a minimization method to minimize a cost function in order to predict the interaction (based on constructed constraints) between these two proteins?

[@SariSabban]

Yes that is possible. Refer to the docking protocol in Rosetta. You can try the demo within the Rosetta package, or try the Rosetta Rosie server

[@nasimsoltani3768]

@@SariSabban Thank you very much. I followed pyrosetta tutorials. Now, I know how to do low resolution and high resolution docking, but I don't understand how I can add my own constraints. Could you please help me with that? Thank you so much

[@sembretunombre9490]

Esta chido, gracias