Foldit design puzzles let players design new proteins from scratch.
One key goal of a design puzzle is to create a protein that will fold up spontaneously the way natural proteins do.
Early efforts to design self-folding proteins in Foldit met with limited success.
The Foldit science team identified "ideal loops" as one way to make player-designed proteins more likely to behave like natural proteins.
Loops are of course sections of a protein that lack a specific secondary structure. Often, relatively short sections of loop connect sections which have a secondary structure of sheet or helix. (In other cases, a long, meandering loop may connect sheets and helixes.)
Short sections of loop often involve a sharp turn. For example, a loop containing only two segments may connect two sheets in an "antiparallel" arrangement.
Analysis of natural proteins has revealed common patterns in these short sections of loop, which has lead to the concept of "ideal loops".
Two scientific papers, Principles for designing ideal protein structures (published in Nature in 2012, and known as Koga & Koga), and Control over overall shape and size in de novo designed proteins (published PNAS in 2015), are the basis of the ideal loops concept in Foldit.
See Design structures for a detailed discussion of the patterns described in Koga & Koga paper.
Several recent changes in Foldit are intended to promote the use of ideal loops in Foldit design.
Ideal loops conditionEdit
Most recent Foldit design puzzles have an "ideal loops" condition (formerly called a "filter"). This condition adds a penalty for any loops which violate ideality standards. The ideal loops condition was introduced in DEVPREV Puzzle 1242 on 3 June 2016.
Blueprint tool Edit
The Blueprint tool was introduced on 27 October 2016.
The Blueprint tool has multiple functions. The main function is in the Building Blocks menu, which shows a selection of ideal loops, with one segment of sheet or helix on each side. Each ideal loop is shown in schematic and three-dimensional form. The 3D form can be rotated to see the effect of the loop on the position of the adjacent secondary structures.
The schematic version of an ideal loop can be dragged to the sequence shown in the Blueprint window. When the schematic is dropped on the sequence, the shape of the protein changes in the main Foldit window. The dropped schematic adds invisible constraints, similar to bands, which tend to hold the loop in its ideal shape.
A dropped schematic can be dragged off the sequence in the Blueprint window. This removes the constraints.
A YouTube video demonstrates the use of the Blueprint tool.
Rama Map and ABEGO coloring Edit
The Rama Map tool displays the Ramachandran plot of the protein. The Ramachandran plot shows the critical phi (f, horizontal x-axis) and psi (?, vertical y-axis) backbone angles of the protein. Although the Rama Map in Foldit does not have an explicit scale, the units are normally degrees, ranging from -180 to 180. In degrees, the lower left-hand corner of the plot is (-180, -180), the upper right-hand corner is (180, 180), and center of the plot is (0, 0).
Both phi and psi are dihedral angles, meaning the angle formed by the intersection of two planes. Both angles are relative to the "alpha carbon" of the residue. See this Proteopedia article on psi and phi angles for a detailed description of how these angles are measured, plus an interactive tutorial. Also Figure 1 of this 2010 Biomolecular Concepts article has a good illustration of phi and psi.
In the Rama Map tool, each black dot represents a segment. A segment can be selected and dragged to a different position, changing the backbone angles for that segment. The Rama Map tool is the first Foldit tool that allows direct manipulation of backbone angles.
The Rama Map tool also shows the "ABEGO" regions as colored areas. When no segments are selected, a generalized ABEGO map is shown. (The generalized map is also shown when multiple segments are selected in the main Foldit window, using the selection interface.) When a single segment is selected, an ABEGO map specific to the amino acid type of that segment is displayed.
Each letter in "ABEGO" refers to a "torsion angle class", which is considered significant in the analysis of proteins.
The letters have the following meanings:
A - "alpha helix" - right-handed helixes, red in ABEGO coloring.
B - right-handed "beta sheet" (or beta strand), blue in ABEGO coloring.
E - left-handed sheets (very rare), yellow in ABEGO coloring
G - left-handed helix (rare), green in ABEGO coloring
O - Cis omega (?) angle, a special case
The cis case represented by the "O" in ABEGO is different than the other four cases. It doesn't involve the phi and psi angles, so it can't be shown on the Rama map. It doesn't have a color in the ABEGO coloring used in Foldit. In the cis case, the bond between two segments is flipped by 180 degrees from the normal position. The normal bond between two segments (known as the trans case) has an omega angle around 180 degrees. A cis bond has an omega angle around 0 degrees.
The effect of a cis bond is that two sidechains in a row will be on the same side of the protein. A cis bond between two amino acids is most common when proline is the second amino acid.
The Rama map shows ABEGO (or at least ABEG) regions. The ABEGO coloring scheme is also through the "AbegoColor" option available under "Color" on the "View" menu. (This option requires that the "Show advanced GUI" option be checked under "General Options".) The sequence shown in the Blueprint tool can optionally be shown in ABEGO colors.
Susume has created two videos showing the use of the Rama map. The first is Foldit Rama Map Series #1 - Fixing a Bad Dot, and demonstrates finding and fixing a segment that lies outside of one of the ABEGO regions. The second video is Foldit Rama Map Series #2- Copying a Loop and demonstrates how to copy an ideal loop from one part of the protein to another.
The Remix function in Foldit has been available for some time, but the user interface has recently been improved. While conceptually similar to the more familiar rebuild function, the remix function may produce results that are more similar to natural proteins. A Foldit blog post by jflat06 (dated 11 May 2016) contrasts remix and rebuild, and describes how to use the new GUI interface.
The new remix user interface is intended to work with the selection interface. The user can select three to nine adjacent segments. No cutpoints are allowed in the remix selection. Clicking the Remix button searches for possible positions in the Remix database. If any valid positions are found, the user can review them one by one, seeing the impact on the overall shape of the protein. Promising poses can be saved to quicksave slots for further development.