This section is designed to guide you through the large number of tools available in X-AUTOFIT, X-POWERFIT, X-BUILD, X-LIGAND and X-SOLVATE.
The aim of this section is to provide a starting point for understanding the X-Ray applications, where the functionality is found, and how the functionality should be used to give the best results.
The X-applications are provided so that X-AUTOFIT, X-BUILD, and X-POWERFIT form an integrated system through a single main palette accessed from the Applications menu on the main menu bar of the QUANTA window. X-LIGAND and X-SOLVATE are separate applications run from the Applications menu.
Electron density bones are integral to the X-Ray applications and provide near automated map analysis and analysis. They are generated from the map only in regions of interest and so are controlled by the map radius. They can be used to generate map masks as well as in the process of Ca-tracing and can be automatically interpreted in X-POWERFIT or semi-automatically in X-AUTOFIT.
Bones are not converted into atoms but represent a background annotation view on which Ca trace atoms are added. Therefore it is not necessary to edit the bones in detail, since the only represent a backdrop guide; let the program auto-edit these with the parameters provided (Bones | Change start value, Bones | Change trim value).
Any editing of the bones is lost by recalculating the bones.
If a map mask is to be calculated from the bones, use a large map radius, and use the bones editing tools to trim the bones to a molecular shaped volume. Use the Map mask | Solvent content tool and the Bones | Bones symmetry tool to check the extent of the mask and bones.
Once a map mask has been generated, it is not necessary to manually trim the bones for any further protocols.
If bones are to be used in X-POWERFIT for automated tracing, use a large map radius, a map mask to delimit the molecular volume (bones/mask bones by mask) and the Auto-trace High (high resolution data) or Find sec. struct. | Auto-trace Low (low resolution data) tools under the X-POWERFIT palette.
If bones are to be used in X-AUTOFIT (and X-POWERFIT) for semi-automated map tracing then use a small map radius (about 9Å is suitable), and work along the trace using the Bones | Next bones box tool to update the view when the trace nears the edge of the calculated volume of bones.
Masks can be generated from coordinate and map information as well as from O format files (old, new and compressed). This functionality is available on the Map Mask palette.
Mask editing is possible by either editing electron density bones before a mask is generated or by directly editing the mask with a spherical pointer.
Masks are used in X-AUTOFIT and X-POWERFIT as bounding masks for all tracing and calculations (Bones | Mask bones by mask) and provide a mask for external solvent flattening functionality in CNX.
Atoms are placed into electron density by the process of Ca tracing using the bones as a background object.
Ca tracing is only possible when the bones are active because bones analysis is integral to creating a Ca trace:
Ca-tracing may be done automatically using X-POWERFIT or in X-AUTOFIT in a semi-automated fashion.
There are three types of Ca-trace atoms: (1) an active Ca atom in (2) an active Ca trace segment, and (3) other Ca-trace segments placed previously.
The active Ca trace atom (and active segment) is set using the Current res/seg tool on the X-POWERFIT, Ca build and Sequence palettes.
If the active Ca trace atom is at a segment terminus, then it can only be moved on the surface of a sphere of radius 3.8Å from the previous Ca trace atom by changing an opening angle and torsion.
If the active Ca trace atom is not at a terminus, then it can be moved in x/y/z screen ordinate. The Ca-Ca distances affected by this are shown on the QUANTA message line.
Only the active Ca trace atom can be moved in one of the following ways:
Only the active segment can be moved or changed by:
The Ca trace is automatically saved between sessions and can be explicitly saved using the CA build | Save changes tool. The last edit of the Ca trace can be undone/redone using the CA-build | Undo last change tool.
A Ca trace is always built from the C-terminal end. Therefore setting the current Ca trace atom as a terminal atom sets the Ca segment polarity so that this atom is the C-terminal atom of a segment.
Selecting a non-terminus Ca trace atom does not affect the polarity of the Ca-trace. The polarity of the Ca trace is shown as an arrow at the current Ca trace atom pointing to the C-terminal (provided that the current Ca trace atom is not a terminus).
Be aware that you must define the polarity of the Ca trace before converting a Ca trace to an all atom model.
If the current Ca atom is at a terminal, the polarity of the Ca trace may be reversed using the CA-build | Reverse chain tool; then the current Ca trace atom is set at the other end of the chain. If the current Ca-trace atom was not at a terminus, the current Ca trace atom is not changed.
The polarity of the Ca trace can be checked with the CA-build | Check CA direction tool, which essentially tests for the C=O position in electron density. This makes no change to the segment direction, only notes whether the current direction is right or wrong.
Sequence assignment is carried out on Ca trace atoms. The residue type is assigned to the Ca trace atom, and, when a sequence is loaded and displayed, the current residue type is shown at each traced Ca atom. The aim is to assign to the Ca trace atom residue types, so the Ca trace is assigned a unique sequence which is part of or all of the loaded template sequence. The loaded sequence represents a guide to this process for the tools provided; once sequence assignment to the Ca trace is complete the loaded sequence template has no further function.
A sequence is loaded (from a number of different formats) using the Sequence | Load sequence tool. If the protein is to be interpreted as a dimer (or higher repeat), it is recommended you load only the sequence for one monomer unit. This way, the application can identify a unique sequence as only one occurrence of the repeat in the sequence.
A Ca trace atom can be assigned one of the 20 amino acids or a number of fuzzy residue types. There are 10 pre-defined fuzzy residue types, and you can add a further 10 fuzzy residue descriptions of your own.
Multiple propensity values can be assigned to a single residue - for example a lysine residue can be classed both as big and small since density for this residue could be either. You are directed to the documentation in Using X-AUTOFIT on specifying new or changing existing fuzzy propensity values.
Picking a residue type sets the residue type of the current active Ca trace atom. Once any Ca trace atom in a Ca-trace segment is set, a sequence alignment is performed.
Sequence alignment is done in a forward and backward direction, and the results are shown against the sequence trace: blue arrows are forward fits, and red arrows are backward fits.
The Ca trace atoms are marked with a unique assignment if a unique sequence is found. The sequence table is updated so that the unique section is shown in upper-case. Any subsequence addition of Ca trace atoms will be added as the correct residue type defined by the sequence.
It is not possible to extend an assigned Ca-trace segment if this results in an overlap of the Sequence assignment or in the sequence alignment falling off the end of the sequence table.
Dimer (or higher order) structures can never be uniquely sequence assigned if the sequence is for the whole protein. Please read in only one monomer sequence unit.
Sequence assignment of a second (or higher order) monomer unit is carried out by first making the sequence assignment in the first monomer "not unique" (Sequence | Unique sequence). This releases the assigned sequence for further assignment.
Sequence assignment is stored with the Ca-trace and does not require the presence of the loaded sequence or a unique setting. On completion of sequence assignment, the loaded sequence can be removed.
Ca-trace atoms are converted to all atom models automatically by one of four tools on the CA build palette. The quality of the results depends on the quality of the placement of the Ca trace atoms as well as the quality of the electron density. It is highly recommended that the actual position of each Ca-trace atom be checked and adjusted in a final refinement (CA build | Refine Current Seg) before trying to build an all atom model.
The all atom model is built with a polarity defined by the current polarity of the Ca trace. The Ca trace polarity should be defined by making the C-terminus atom the current atom (CA build | Current res and seg). This can be checked against the sequence by examining the arrows and by using the CA build | Check CA direction tool to check against the map.
The electron density map must cover all the Ca-trace atoms if fitting is to be carried out to electron density. Change the map resolution so that all Ca trace atoms are covered. Turn off the map during the building process, since the repeated re-display of this during the building slows the calculation down.
Knowing which of the four tools to use depends on the quality of the electron density. You are strongly advised not to use the CA build | Fit seg by database tool, since it produces poor results.
The new coordinates produced by the building process will either be put in a new MSF file or placed in a current MSF file:
If a MSF file is open/active and visible, then the new coordinates are added to the first active and visible MSF file.
If there is no open/active and visible, MSF molecule then a new MSF is generated.
Therefore, if you do not want to add new coordinates to a currently open molecule, then ensure that the molecule is not active.
Model building is carried out using the following palettes from the X-BUILD application: Build atoms, Structure, Pointer, and 3D text. Additionally, validation tools are used: data logging with the Last commands tool and advanced validation using the Table/graphs palette.
Atom selection for building is carried out using:
To edit a single residue, use the tools on the Build atoms palette, and use the Structure palette to refine of a region/volume of residues.
The Pointer palette (generally visible all the time) and the 3D text palette should be used to center the view.
X-BUILD has three types of real space torsion angle refinement protocols: gradient, grid and Monte Carlo. In all cases you should be aware that if the electron density does not cover the atoms of interest, then the functionality does not produce correct results.
All the refinement tools improve fit to density by modifying torsion angles. The grid refinement tools also vary angles slightly to improve the radius of convergence, and this can be corrected subsequently using geometry refinement tools. The gradient refinement tools allow small variation in bonds and angles, although these are generally small and do not need to be corrected by a model building tool.
Grid refinement tests all possible conformations by changing a small number of torsion angles to find the best fit to electron density. It is therefore limited to side chain and main chain protein and nucleic acid fitting.
Grid refinement is used to place side chain atoms or main chain peptide planes into an electron density during the residue by residue fitting process. The tools Build-atoms | Fit side chain by RSR, Build-atoms | Move atoms + RSR and Build-atoms | Fit main chain by RSR carry out Grid refinement.
Grid refinement can only be used on proteins and nucleic acids.
Grid refinement places the sidechain atoms into an electron density at some detriment to angles and improper angles. Hence, it will produce a slightly distorted sidechain, especially where the electron density is poor. It is therefore necessary to regularize residues after grid refinement.
Gradient refinement is used to refine any residue or group of residues into an electron density by following the electron density gradient. A single residue can be fitted to an electron density using the Build atoms | Refine 1 residue tool, and multiple residues are fitted using the Structure | Refine zone tool. These tools use torsion angle refinement, so any defined torsion is allowed to change.
Gradient refinement can be used on any residue or group of residues (protein, nucleic acid, water, ligands).
The refine 1 residue tool does not improve the geometry but only refines toward a better electron density. If the bonds/angles and improper angles were initially bad before the use of this tool, they will be bad after using the tool as well.
The Refine zone tool improves the fit to electron density and improves geometry terms using mixed parameter refinement. Hence it is possible that the fit to electron density may be worse after using this tool if the geometry was initially poor.
The Structure | Rigid body fit tool allows rigid body refinement of regions of residues. This is generally useful to improve the fit of an entire domain after MR.
Monte Carlo refinement is used in X-BUILD to place main chain conformations of loops and termini where there are too many degrees of freedom and the atoms are too distant or the density too poor to use gradient refinement.
Structure | Loop fit and Structure | Terminal fit are used to fit loops and termini respectively. In both cases they should be considered a "try it and see" protocol where the results take a number of minutes to calculate but require no user input.
Geometry refinement is used to improve the bonds/angles and improper geometric terms of atoms. It takes no account of the electron density. This is known as regularization and should be used after side chain fitting and single residue refinement, or after a manual modeling session.
The following tools all carry out geometry refinement:
Build atoms | Move atom + reg. res.
Build atoms | Move atom + reg. zone
Structure | Refine Zone also carries out geometry refinement as part of the electron density refinement protocol.
We suggest that proteins and nucleic acids be fitted by traversing the polypeptide chain one residue at a time using the Pointer | Next residue tool. If the residue does not fit the experimental data, then use the Build-atoms | Fit side chain by RSR or Build atoms | Move atom + RSR tools. The residue should be tidied up with the Build-atoms | Regularize tool.
If a number of consecutive residues have problems, use the refine zone tool on the Structure palette or the Build-atoms | Move atom + reg. res. tool.
Manually editing residues is, in most cases, unnecessary.
Ligands can be added in the application X-LIGAND, and waters are added en masse in X-SOLVATE. A single or a small number of waters can be added using the Build-atoms | Add-delete | Add water at pointer tool followed by gradient refinement with Build-atoms | Refine 1 residue.
Ligands are automatically parameterized for both real space torsion angle refinement and geometry refinement. Hence it is possible to edit these without generating complex molecular descriptions. Generally, ligands should be fitted to electron density (if not already placed in X-LIGAND) with the Build-atoms | Refine 1 residue tool.
All waters can be refined in a single step with the Structure | Refine all water tool; it is also possible to inspect individual water refinements when there is a problem in the Structure | Do all... water refinement option.
Use the Ramachandran plot to identify regions of backbone chain that need rebuilding. The Ramachandran plot can be picked to place the view at a residue in an unlikely conformation.
Alternate conformations can be automatically searched and added from the Structure | Do all... tool, but this does require that the structure is near the end point of refinement.
Only the first active and visible molecule can be edited. Any molecule edited will be automatically saved on exit.
Use the 3D text editor to mark problems in model building. These can then be checked in subsequent model building using further refined coordinates.
Use the Validate tool on the main X-BUILD palette to identify problems remaining in the structure at the end of a model building session. Any problem (except a Ramachandran error) can be automatically fixed with the 3D text | Fix validate error tool.
When the structure is near the end point of refinement use the advanced validation tools on the Table/graphs palette to identify more subtle problems with the structure.