Bug in secondary structure filter
| Case number: | 671071-2012042 |
| Topic: | Game: Other |
| Opened by: | ichwilldiesennamen |
| Status: | Open |
| Type: | Bug |
| Opened on: | Friday, September 10, 2021 - 07:00 |
| Last modified: | Tuesday, September 14, 2021 - 06:30 |
I tried to do a helix-only design on puzzle 2041 and was surprised to see that I got penalized for 122 helix-segments even though there are only 80 residues in the puzzle. Seems like the number of helix-segments is doubled in the filter.
Because if I add one helix-residue to a design which has just enough loop/sheet segments, then I get penalized for 2 helix-segments.
This should be easy to fix because it seems that you just have to divide by 2 somewhere.
Oh ok, there is indeed a 4x penalty for one additional helix-segment on the monomer. See my shares iw6a and iw6b on this (b has 1 helix-segment more on the monomer). But still, in iw6a I got penalized for only 2 helix segments. Seems like there is an offset of 2 penalty-segments.
With this high 4x-penalty it does indeed not make sense to me to try a helix-only design so I will refrain from this.
Since puzzle 2041 is a symmetric tetramer puzzle with 4 symmetric chains, I believe per-residue penalties are applied to all residues in the puzzle (not just the monomer subunit). However, I am surprised that you would ever see a penalty for just 2 helix residues. If you have a solution like that, could you share it with scientists?