The contact map also shows black dots when segments are actually in contact. Generally, two segments come into contact when they're less than 8 angstroms (8 band length units) apart. Different pairs of amino acids have different contact distances.
The contact map is available on certain Foldit puzzles, which generally have "predicted contacts" in their titles.
The contact map is displayed by the "Contact Map" button on the Actions menu in the original interface, or the Main menu in the selection interface. There is no keyboard shortcut for the contact map.
The contact map is a floating window, which can be moved around in the main Foldit window. Dragging on the rows of dots at the bottom of the contact map changes the size of the window.
Reading the Contact Map
The contact map is complex, and can display several different types of information. The most prominent feature is the diagonal line running from the lower left to the upper right corner, which represents the backbone of the protein.
The two sides of the backbone diagonal are mirror images of each other. Both sides contain the same information.
The background of the contact map is a grey-and-white checkerboard. Each square or "cell" represents a pair of segments. (The checkerboard pattern is not always obvious, depending on the computer's display settings.)
Cells colored in shades of green are predicted contacts between two segments. A black dot in a cell indicates the corresponding segments are actually in contact, regardless of whether the contact was predicted.
Hovering over one of the cells highlights the row and column, and displays the contact information at the top of the contact map window. Also, hovering on a cell in the contact map draws a "candy cane" spiral between the corresponding segments in the main window. The spiral is colored in shades of green if the segments are in contact, and shades of red if they aren't.
The contact information is the pair of segments and the "weight" of the contact. For example, the pair of segments might be 110,8 and the weight 1.0. (The contact map always shows the higher-numbered segment first when it displays the contact information.)
Contact weight has (also called "heat") is the probability that two segments are in contact. In theory, weight could range from 0.0 to 1.0, but in practice, the lowest probabilities are rounded down to zero. In a typical contact map, many cells have a probability of 0.0, while others range from around 0.3 to 1.0.
Lower contact weights are represented by darker shades of green in the contact map.
The backbone diagonal
The diagonal is a special case. The diagonal represents the backbone, so it shows [ 1, 1 ], [ 2, 2 ], [ 3, 3 ], and so on. A segment is always in contact with itself, so Foldit uses these cells to display the current secondary structure, with blue cells for helixes and orange cells for sheets. The diagonal never has any predicted or actual contacts.
Using the Contact Map
Hovering on a contact map cell displays the contact information and red or green contact spirals. This information disappears when the mouse pointer is moved off of the cell.
Clicking on a cell in the contact map selects the corresponding contact. The cell is highlighted with a black rectangle, and the contact lines in the main window remain until the cell is deselected. Clicking on a cell again deselects the contact. Clicking on a contact line in the main window also deselects the corresponding cell in the contact map.
Multiple cells can be selected. Even cells without predicted contacts can be selected.
Clicking on a cell in the backbone diagonal has no effect.
The contact map has four tools for working with selections:
- Clear Selected Cells - de-selects all cells of the contact map
- Band Selected Cells - creates a band between any selected cells on the contact map
- Select Contacts - selects all contact map cells with a weight greater than zero
- Select Empty Contacts - like Select Contacts, but only selects cells which aren't currently in contact
Interpreting the Contact Map
The patterns of contacts in a contact map give clues about the shape of the protein.
The easiest clue is when there is a group of contacts close to the backbone diagonal. This tends to indicate the section of backbone should be a helix. Changing the secondary structure to helix and applying the Idealize Secondary Structure tool should increase the number of segments in contact for this section.
Another clue appears when there is a group of contacts parallel to the backbone diagonal. This tends to indicate that the two sections of the protein are parallel to each other in space, and also parallel in segment numbers. So the highest-numbered segment in one section is closest to the highest-numbered segment in the other section.
Just knowing that two sections are parallel doesn't immediately tell you whether they are sheets or helixes. In general, the contacts for sheets should be a thin line, and the contacts for helixes should be thicker.
Yet another clue appears when a group of contacts is at a right angle to the backbone diagonal. This indicates the two sections are anti-parallel in terms of segments, so the lowest segment of one section is closest to the highest segment of the other section.
Anti-parallel sections can still be parallel to each other in space. This is especially common with sheets, where one sheet is often followed by a few loop segments which bend in tight turn of around 180 degrees. Following the loops is another sheet, which forms hydrogen bonds with the first sheet. The two sheets in this scenario are anti-parallel in terms of segment numbers, but they're parallel in space. Anti-parallel sheets generally form stronger bonds than parallel sheets.
Contact maps and distance maps
Contact maps are a common scientific tool, and are closely related to "distance maps", which show the actual distances of segments in a solved protein.
The contact maps in jeff101's page simply indicate whether two segments are within a certain distance of each other. They're based on the actual shape of solved proteins. In Foldit, the contact maps contains predictions of where contacts may occur.
Contact Map numbering
Understanding how the cells on the contact map are numbered can be helpful when looking at the protein in the main window.
As mentioned above, each cell on the contact map represents a pair of segments on the protein, with the diagonal representing the backbone. The cell in the lower left corner is [ 1, 1 ]. If the protein has 85 segments, the cell in upper right is [ 85, 85 ].
Hovering over a segment of the protein in the main window highlights the segment's backbone cell on the contact map, and the row and column. For example, if you hover over segment 94, cell [ 94, 94 ] on the contact map is highlighted.
In the contact map, hovering on the same spot on the diagonal and moving horizontally, the first number remains the same, and the second number decreases as you move away from the diagonal. So you'd see [ 94, 93 ], [ 94, 92 ], [ 94, 91 ] and so on moving horizontally.
Moving vertically from the diagonal, the second number remains the same, and the first number increases, so you'd see [ 95, 94 ], [ 96, 94 ], [ 97, 94 ] and so on.
The farther you move from the backbone diagonal, the more distant the contact points are. The upper left and lower right corners of the contact map represent contact between the first and last segments of the protein. Both these cells represent [ 85, 1 ] in an 85-segment protein.