tommyhuangthu
commented
8 months ago Hello Gabriel,
I am very sorry for the late response and thank you very much for your interest. Because UniDesign employs a simulate annealing Monte Carlo-based stochastic searching strategy, the solutions may have some variability. However, if you always get the same solution, it means that the algorithm may converge and find the global energy minimum for your system. This can happen when you have too few designable residues in your RESFILE.txt. You can increase the number of mutable and repackable sites to gain more variability.
If possible, you can share your .pdb and RESFILE.txt files with me so that I can have a look for you.
Thank you!
Best wishes, Xiaoqiang
From: Gabriel Ernesto Jara @.> Sent: Monday, January 8, 2024 9:54 To: tommyhuangthu/UniDesign @.> Cc: Subscribed @.***> Subject: [tommyhuangthu/UniDesign] Sequence variability when preforming a ProteinDesign on a protein-protein interface (Issue #1)
Hi! Thank you for making available your program, it is very useful.
I have some questions about the design of protein-protein interface. I run the next command: ./UniDesign --command=ProteinDesign --ppint --pdb=complex.pdb --design_chains=DE --resfile=RESFILE.txt --ntraj=10. I expected to obtain 10 different solutions (sequences), however, all of them converged to the same (unique) sequence. I tested to increase the trajectories values to 100, but it did not change anything. Is this an expected behavior? Is there anything I can do to get some variability in the design sequence?
Best
Gabriel
— Reply to this email directly, view it on GitHubhttps://github.com/tommyhuangthu/UniDesign/issues/1, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AIUXW3CTCXE3E2JYZED54J3YNQCAXAVCNFSM6AAAAABBRSUCB6VHI2DSMVQWIX3LMV43ASLTON2WKOZSGA3TANRSGQYDOMY. You are receiving this because you are subscribed to this thread.Message ID: @.***>
Hi! Thank you for making available your program, it is very useful.
I have some questions about the design of protein-protein interface. I run the next command: ./UniDesign --command=ProteinDesign --ppint --pdb=complex.pdb --design_chains=DE --resfile=RESFILE.txt --ntraj=10. I expected to obtain 10 different solutions (sequences), however, all of them converged to the same (unique) sequence. I tested to increase the trajectories values to 100, but it did not change anything. Is this an expected behavior? Is there anything I can do to get some variability in the design sequence?
Best
Gabriel