RosettaCommons
Cyclic peptide design issues
Hello, I'm a beginner and would like to use RFpeptides to design cyclic peptides that bind to my target protein. I've generated over 2,000 structures for my target protein, but none of them have resulted in a cyclic peptide. I'm worried it might be a script issue. I used the official sample input PDB and the script twice to generate a total of 20 structures, but still haven't found a cyclic peptide. Could you please help me check the following to see if this is a script issue?
#!/bin/bash
prefix=/data/*/job1/output/new_cyclic_binder
num_designs=10
pdb='/data/*job1/7zkr_GABARAP.pdb'
HOTSPOTS='["A51","A52","A50","A48","A62","A65"]'
script="/data/RFdiffusion/scripts/run_inference.py"
$script --config-name base
inference.output_prefix=$prefix
inference.num_designs=$num_designs
'contigmap.contigs=[12-18 A3-117/0]'
inference.input_pdb=$pdb
+inference.cyclic=True
diffuser.T=50
+inference.cyc_chains='a'
+ppi.hotspot_res=$HOTSPOTS
This issue discusses the same problem. It seems that for some people reconfiguring the SE3 environment helped.
Thanks a lot.I will try it.
发件人: Rachel Clune @.> 发送时间: 2025年9月30日 0:20 收件人: RosettaCommons/RFdiffusion @.> 抄送: zzxchelly @.>; Author @.> 主题: Re: [RosettaCommons/RFdiffusion] Cyclic peptide design issues (Issue #406)
[https://avatars.githubusercontent.com/u/34350601?s=20&v=4]rclune left a comment (RosettaCommons/RFdiffusion#406)https://github.com/RosettaCommons/RFdiffusion/issues/406#issuecomment-3347929927
This issuehttps://github.com/RosettaCommons/RFdiffusion/issues/359 discusses the same problem. It seems that for some people reconfiguring the SE3 environment helped.
― Reply to this email directly, view it on GitHubhttps://github.com/RosettaCommons/RFdiffusion/issues/406#issuecomment-3347929927, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AH4T54JQGQIBVDFND7Q6NRL3VFL5FAVCNFSM6AAAAACHU7K3OGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGNBXHEZDSOJSG4. You are receiving this because you authored the thread.Message ID: @.***>