The run_inference.py section hasn't been changed

[hzp13151231131]

The run_inference.py section hasn't been updated, and design_macrocyclic_binder.sh cannot design cyclic peptides.

[davidcjuergens]

Hello, I don't believe run_inference.py needs to change in order to be able to run the cyclic peptide design. There are other modifications that enable this. Can you please be more specific with the issues you are facing? Are you seeing errors? Are the peptides simply not coming out cyclized?

[hzp13151231131]

Hello, I don't believe run_inference.py needs to change in order to be able to run the cyclic peptide design. There are other modifications that enable this. Can you please be more specific with the issues you are facing? Are you seeing errors? Are the peptides simply not coming out cyclized?

Thank you for your response. I have run tests on the design_macrocyclic_monomer.sh script using the case provided in the paper, aiming to de novo generate macrocyclic peptides from 7zkr_GABARAP.pdb. However, the resulting peptide sequences did not form cyclic structures and remained linear, even after multiple attempts. I look forward to your guidance on this issue.

[nalaterry]

Yes, I also encountered the same problem. Only linear peptides could be obtained instead of cyclic peptides.

[Sandippanja29]

Same thing is happening for me also , after using the two new flags inference.cyclic=True and inference.cyc_chains='a' still it is only generating structured peptides instead of cyclic peptides.

[davidcjuergens]

Hmm, interesting. This seems like a significant and reproducible issue--thanks all for noting your anecdotes. I will look into this now and report back, I apologize for the hiccup everyone!

[jkamishi]

I apologize if I misunderstood. After opening the output structure in PyMOL, display the sequence, select the peptide sequence, and try showing it as a “lines” or “sticks”. It may be displayed as a cyclized.

[davidcjuergens]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

[nalaterry]

Thanks for your reply. I followed your suggestion and confirmed that it's not a visualization issue. I've tested the "examples/design_macrocyclic_binder.sh" command multiple times, and the resulting binder is consistently a helical structure. I'm using the latest version of the GitHub repository.

[hzp13151231131]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

Thanks for your reply.Of course，I try to open the output structure in PyMOL, display the sequence, select the peptide sequence, and try showing it as a “lines” or “sticks”.But it also shows linear peptide.

[Sandippanja29]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

[nalaterry]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

Hello, are you using this GitHub repository to run RFpeptide, and is the command "examples/design_macrocyclic_binder.sh"?

[nalaterry]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

Hello, are you using this GitHub repository to run RFpeptide, and is the command "examples/design_macrocyclic_binder.sh"?

Something interesting happened! I reconfigured my SE3nv virtual environment and was then able to design cyclic peptides.

[hzp13151231131]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

Hello, are you using this GitHub repository to run RFpeptide, and is the command "examples/design_macrocyclic_binder.sh"?

Something interesting happened! I reconfigured my SE3nv virtual environment and was then able to design cyclic peptides.

Thank you！I will try this！

[Alberto-tycho]

hi @davidcjuergens thank you for your support can you post the link to the Stephen Rettie Google Colab notebook that works for cyclic peptide designs? thank you in advance!

[DazLe-Q]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

Hello, are you using this GitHub repository to run RFpeptide, and is the command "examples/design_macrocyclic_binder.sh"?

Something interesting happened! I reconfigured my SE3nv virtual environment and was then able to design cyclic peptides.

Hello, how can you reconfigure the environment? I am encoutering the same issue as you.

[nalaterry]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

In my case i can see the peptides as circular now! Thanks @davidcjuergens

Hello, are you using this GitHub repository to run RFpeptide, and is the command "examples/design_macrocyclic_binder.sh"?

Something interesting happened! I reconfigured my SE3nv virtual environment and was then able to design cyclic peptides.

Hello, how can you reconfigure the environment? I am encoutering the same issue as you.

SE3nv.txt Here are the packages in my conda environment. Hope this helps. @DazLe-Q

[rclune]

yes the code is working when attempting to run from a Google Colab notebook that Stephen Rettie set up. indeed the cyclic nature of the peptides is more obvious when you show sticks and util.chainbow in pymol. @nalaterry @Sandippanja29 @hzp13151231131 can you try this?

Is the CoLab notebook you mentioned publicly available?