The ‘chains to scan’ are a subset of the 'chains for surface evaluation': For these chains the surface is determined AND they are scanned. Only for these chains a peptide-library is generated.

Why is it necessary to distinguish between chains for surface evaluations and scanning?

Imagine you have the crystal structure of a protein, containing the chains A and B.

Case 1: It is known that the chain A as monomer is binding the ligand; you will select only this chain for surface evaluation as well as for scanning.

Chain A

surface

Chain A

surface + scanning

Case 2: It is known that the ligand binds to chain A AND B as dimer; it is necessary to choose both chains for surface evaluation as well as for scanning.

Chain A and B

surface

Chain A and B

surface + scanning

Case 3: It is known that the monomer A as well as the dimer A+B (and not monomer B) bind the ligand, implying that the interface from chain A to chain B is not of interest for the binding. If chain A would be used for surface evaluation and for scanning, this interface would be contained in the scanned sequence; if chains A+B would be chosen for surface evaluation and scanning than the surface of chain B would also be scanned. Choose chain A+B for surface calculation and only chain A for scanning, and only these parts of chain A are scanned that are accessible also in the dimer A+B.

Chain A and B

surface

Chain A and B surface +

only chain A for scanning

REGIONS:

Why to define a region?

If you have no information about the protein, than it is not necessary to choose a special region. When you start the program a first region containing the whole sequence is already predefined.

If the number of peptides you want to generate is limited and if you have special information about the binding site, it is possible to restrict the area to be scanned. It is possible to define several regions.

Evaluated peptides must be fully contained in one defined region.

How to define a region?

There are two possibilities to define a region; either by selections in the sequence window (1) or by selections in the "show"-window (2).

1. Sequence window

If you already know the amino acids defining your region then define your region by construction with marked sequence parts: Open the selection window (strict beyond the help window, on the right). If you want to define a new region choose the amino acids in the sequence window, and press the "make new" button - the selection becomes your new region.

If you want to edit an existing region, choose this region as active region (in the left near the selection window), mark sequence parts which should be added or removed, and press the corresponding button.

Here an enlargement of an already edited region is possible.

2. Show window

a) You know the location of the region on the protein surface (for instance if you have a crystal structure with another ligand, binding in the same active side) you should first rotate the protein so that the region can good be separated; move the clipping planes in the side view-window so that the whole region is contained between the planes, and parts that are unimportant are outside the planes. If the clipping planes are not visible, zoom out by pressing ALT and moving the mouse over the show window with pressed left mouse button. Click in the show window on an atom near the centre of the wished region. If you missed the atom 'no atom' is written in the field left of the "save as povray"-button over the side view window, otherwise the name of the atom is written in this field. When the atom is chosen, click on the "define region"-button. A new window appears were you can choose your parameters. Move the window such that your region is visible. Now you can either edit an existing region, or create a new one.

Now you can determine the region’s shape and size (first and second radius). The corresponding shape is simultaneously shown in the window below and the appertaining amino acid in the sequence window. Change the parameters until the region meets the requirements and press “add region”. Now a new region is created, which is exactly defined by the shape and the clipping planes.

b) An already existing region should be edited:

Choose the region that should be cut. You see an sequence window were the first two lines show the whole sequence of the scanned chains, and the third line shows your active region. The 4th line contains only those amino acids which are part of your currently active region AND the new shape.

Note: Within the show-window, an active region is an already defined region and therefore can only be shortened not enlarged.

Press “update” to replace the active region with the newly edited (= shortened) region.

c) If you want to define the whole sequence as a new region, select the “new maximal region.

I want to define a region but the ellipsis or circle is not visible. What to do?

The circle is not part of the window where the protein can be seen; it is only painted on this window, and vanishes when other windows are moved over. If this happens, change the parameters a little bit (e.g. change between circle and ellipsis and back), and it is visible again.

Please move the region window so that the window does not cover your chosen region in the show-window.

How (to what extend) does the “ProteinOptions” affect the peptide evaluation?

When the chains for surface determination are chosen and the (Connolly) surface is evaluated, all atoms of the amino acids are classified as accessible (=superficial) or not-accessible (=not-superficial). Since peptides consist of whole amino acids and not only atoms, it has to be defined whether the whole amino acid is considered as accessible or not.

There are two extreme examples to clarify the necessity of those definitions:

1) If you claim that every atom of an amino acid should be accessible you will get only a very few accessible amino acids.

2) If you want that at least one atom of the amino acid should be accessible, you will get many amino acids as accessible. Meaning that one accessible atom is sufficient to define the whole amino acid as superficial even when most of the atoms are buried.

You have to find a trade-off depending on the proteins structure:

Define, as a threshold, a special percentage, e.g. 50%, of the atoms of the amino acid to be accessible to define the whole amino acid as accessible. The percentage can depend of the protein structure: it will be larger for well-packed crystal structures, and smaller for loosely packed NMR structures or models of proteins.

Now the surface consists of different sized sequences of accessible amino acids, interrupted by non-accessible amino acids. For the scanning and the generation of peptides these parts should have a minimal length. This might be a problem: E. g. if you look at superficial helical parts of the protein, you will get two accessible amino acids, than one or two non-accessible, than again two accessible, and so on. You will not get a continuous sequence of accessible amino acids. If you scan only the superficial amino acids such a helical part will be lost, because one or two amino acids are too short to form a peptide. Superficial will simply reject them. Therefore it is essential to fill up those small gaps between accessible amino acids and, vice versa, delete short accessible parts of the sequence between two large non-accessible parts.

The parameters for this algorithm are "window size" and "percentage of amino acids at surface in the window". A user-defined window scrolls over the sequence. For this window the user defines a minimal percentage of accessible amino acids. A gap between accessible amino acids is closed, if this percentage is reached.

For example:

An gap of length one is to be closed: choose window size of 3 and an percentage of 66.

An gap of length two shall be closed: choose window size of 4 and an percentage of 50.

An gap of length two is to be closed, but only in cases were not further gaps are situated close-by on both sides of the gap:

Choose for instance a window size of 6 and percentage 66.

(xx__xx will be closed, x__xxx will be closed, xxx_x__x_xxx becomes xxxxx__xxxxx – median gap is not closed)

PEPTIDES

What does the minimal and maximal length of linear peptides mean?

The aim is to synthesize peptides which are not longer than the user-defined maximal peptide length, but do not miss shorter peptides with at least the user-defined minimal length.

1) Peptides which are shorter than the minimal length, will be rejected

2) For peptides with an intermediate length (between minimal and maximal length), the whole sequence will be taken.

3) For peptides larger than the maximal length you will get several shifted peptides, each of them with the maximal length.

What is the intention of the minimal strand length and maximal diameter of non-linear peptides?

In contrast to linear peptides, non-linear peptides consist of several spatial neighbouring strands (= parts of the original sequence). To discover such strands, a virtual sphere of a given diameter is constructed. It is up to the user to define the minimal length of such a strand as well as the maximal diameter if strands are joined. The strands inside the virtual sphere are combined and connected by a linker in such a way that the whole sequence will be as short as possible.

So strands have to meet several conditions: Each strand should have at least the minimal length, every strand should be inside the sphere, inside the region, and part of the surface.

Please remember that the distance between two neighbouring C_αs is about 4 Å, hence the diameter of the sphere should be at least 4 times the minimal strand length; And keep in mind that the resulting non-linear peptides can be longer than the linear peptides. You can get several strands of the maximum length plus several linkers and possibly end with peptides up to 30 amino acids.

If the found strands are part of a parallel beta-sheet, and you have 3 beta strands of length e.g. 5 found in the constructed sphere, the necessary linkers will become at least 6 or larger, and therefore the whole peptide between 25 and 30 amino acids; but the same situation with antiparallel strands will give a peptide between 15 and 20 because of the smaller linkers.

The program evaluates only linear parts of the sequence, and no non-linear ones. What went wrong?

There are two possibilities: Either the parameter "max. peptide diameter" is too small (all parts of a non-linear peptide must be contained in a sphere with this diameter), or the defined regions are too small.

REPLACEMENT OF AMINO ACIDS

What is the difference between the options "replace all amino acids" and "replace buried amino acids", and how to change the values?

There are amino acids that are avoided in peptide synthesis. For instance cysteines in peptides are often replaced by serine. You can automatically replace such amino acids. Write the one letter code of the amino acid which should be replaced (here C) in the first line, the code of the second (here S) in the second line, and press update - the table of the replacements will be updated. If you made a mistake, and you want to reverse, then replace the amino acid by the same (e.g. replace C with C), and the wrong entry vanishes.

If necessary you can distinguish between buried (non-accessible) amino acids and superficial ones

Example:

At the surface of a protein there are two cysteines. For the first one e.g. 80% of the amino acids are superficial whereas for the second just 60% are superficial. The first cysteine should build a disulfide bridge with a ligand and the second one builds a disulfide bridge to another cysteine within the protein. You can set a threshold to define the second cysteine as buried. Hence the buried cysteine will be replaced with serine but not the first one.

FILES

Why is it necessary to choose a path for protein files?

It is not essential to choose this path, but it speeds up the work. Once chosen, this path is always opened if you want to load a new protein.

Why is it necessary to choose a path for surface files?

The evaluation of the surface of a protein can take time, especially for large proteins. To avoid unnecessary and repeated evaluations, the surface-files are saved and reused if possible. By the way, the program recognizes if the surface files are valid for the protein, so it is no problem to work with different models of a protein with the same file name.

My existing saved project cannot be reloaded.

The project file contains most of the information to reconstruct the project but it doesn't contain the whole pdb-file but the original path to the file and the file name. It additionally contains the path of the correspondent surface files but not the surface information itself.

Please don’t load the protein file via network, or from paths that change from user to user, if a different user should open the project.

Please do not change the name or the path of the protein file while they are used by Superficial.

If you need to reanimate such a project please try to reconstruct the situation (file name, path) when the project was saved.

During the surface evaluation the program crashes. What happened?

There are few proteins with degenerated arrangements of some atoms stored in the PDB. The log-file of the program contains the numbers of these atoms. Try to avoid these atoms by choosing other chains, or change the co-ordinates of one of these atoms a little bit...