Puzzle 2061 (Display + Objective Penalty)

Started by Vincera

Vincera Lv 1

I am experiencing 230pt penalty for too many residues in helices - but my solution is ALL sheets; no helices.
Also: the protein is dimmed (tried all Fandom tips and shortcuts) [not a background display issue].

LociOiling Lv 1

For the too many helix residues issue, first try the auto structures tool to confirm your sheets are really sheets.

Use Actions -> Auto Structures in the original interface, or select all -> L -> Auto in the selection interface.

Manually assigned structure doesn't matter for the penalty, it's only what auto structures thinks that counts.

If there's still an issue, do a share with scientists, and mention the problem in the description.

The message you get is a little wonky. I tried making one big helix, and the message says I have 94 too many residues in helices. Nice trick when the puzzle has only 70 residues. There are 210 total residues if you count all three chains. So it would almost make sense if it said there were 105 too many segments in helices. Not so sure about the 94, maybe you get some bonus helix segments.

On the dim protein issue, that can happen when the chains are far apart. Try alt + shift and dragging upward on the background, which should make the far stuff easier to see. That's the "fog" control.

There's also control + shift and drag, which is "far visibility", and more of an abrupt cutoff compared to the fog. And there's control + alt and drag for "near visibility".

These three visibility controls sometimes pick up weird settings. The "q" shortcut is one way of resetting them, and the Home key is another.

The "q" shortcut is generally good at putting everything onscreen when things are far apart. Far away parts may still be dim in the fog. With "q", the protein gets re-centered, but zoom level is left alone. That's great unless you're zoomed to an extreme level.

Using "Home" both re-centers everything and resets the zoom level. If you have chains far apart in a symmetry puzzle, you may be looking at empty space this way.

The recipe FindIt uses bands to help locate lost chains on a symmetry puzzle.

I'm not sure if the visibility controls can fix all cases where the protein is dimmed or cut off unexpectedly in a symmetry puzzle. But they're a place to start. Usually, the odd behavior goes away when the chains are reasonably close together. But like an electron density puzzle, the display can change in ways you don't expect when you rotate the protein.

Vincera Lv 1

As ever, I thank you, Loci. I was up until 04h30 with this buggy business on several different submissions. I shall pursue further your suggested resolutions.

LociOiling Lv 1

The excess helix penalty is a little weirder than I thought.

It seems you get 36 helix segments with no penalty. With 37 segments in helix, the penalty is for one excess segment. After that, for every extra helix segment, the penalty is applied for three segments.

The results look like this:

helix segments excess helix
36 0
37 1
38 4
39 7
40 10
41 13
42 16

Vincera Lv 1

Ditto that. One of my downloaded AF predictons on a 4H (99SIM) generated a lesser penalty than a 8E (all sheets) solution. I compared several and shall submit my digital lab notes to the Scientists as you suggested earlier.

Vincera Lv 1

All of my TIM barrel and beta-propeller and "beta-buckies" experiments were for naught on 2061. And the SMD hotfix came out too late for me to go back to that one. A dismaying week - and it's only Tuesday. :(