Assignment : Quality Reports

Analyses of Potential Errors and Completeness in Assignment

This popup presents a series of tables that enables the user to get a sense of how complete the assignment is within a given project and whether there are any assignments which look like they might be mistakes. This system is not designed to say that something is definitely missing or definitely a mistake, but rather tries to focus attention to any places where there may potentially be a problem. Often something that is highlighted in red as “unusual” merely needs checking, to ensure that the user has a good explanation. The layout consists of four tabs, each relating to a different aspect of assignment.

Atom Assign Status

The first tab contains various categories of atom, from the selected molecular system, like “backbone” or “Lys” (for amino acid type lysine). Each category has an indication of how many, out of the total available for that type, have been assigned and what chemical shift range they cover. This kind of information is useful to track the assignment progress of a project and to give a final summary. By default the table represents the atoms in the whole of the selected molecular system, but the user may specify only a subset by typing the appropriate chain letters and residue ranges into the “Residue Selection” field. Only atoms with element type “H”, “C”, “N”, “P”, “F” are considered.

Explanation of categories, in terms of atom name:

  • (Backbone)N+H: “H” or “N”
  • Backbone+H+HA: “H”, “HA”, “HA2”, “HA3”, “C”, “CA” or “N”
  • Backbone: “C”, “CA” or “N”
  • Side Chain = not (Backbone+H+HA)

Through-Space Status

The second tab contains a table that gives the counts of various classes of through-space connection, e.g. for NOESY experiments. Counts are given on a per-chain and per-residue basis. In this way the user can gauge the information density that has been or may be used during structure calculation. By default the table shows data for all peak lists that have and experiment type that is deemed to be “through-space”, like NOESY or solid-state dipole-dipole. Data may be displayed for only one of these peak lists by selecting from the “Peak List” pulldown menu.

Resonances

The third table considers all of the resonances that carry chemical shifts in a specified shift list. The purpose of this table is to indicate potential problems with assignment from the perspective of a resonance, rather than peaks, which are covered in the next tab. Three basic category of error are highlighted (and appear as red cells):

  • Each resonance is checked for which others it has covalent bond type connectivity to, considering both atom assignments and the peak dimensions it is assigned to. Typical errors include hydrogens with more than one bound partner and carbons with more than four. Often errors are the result or (redundant) resonance duplication.
  • The chemical shift of the resonance is compared to the known shift distribution derived from BioMagResBank data. Outliers are coloured yellow, orange or red in the “Type Score” column according to how unusual the chemical shift is. Even red warnings may actually represent correct atomic assignments, but the user should be able to conform the assignment.
  • The distribution of peak dimension position information, from which the chemical shifts are derived, is used to fill the “SD” (standard deviation) and “Max Peak U+0394” columns. The former gives and indication of the width of spread in the shift measurement, and is weighted by spectrum dimension, like all shift measurements in Analysis. The later represents the biggest difference between the chemical shift value (an average) and peak dimension position; and thus is useful for finding outliers, especially where the number of measurements is large and the standard deviation is small.

The user may eliminate rows from the table which do not carry any warnings by selecting the “Show only alerts” option.

Peaks

For a selected peak list, the forth table displays assignment and intensity problems on a per-peak basis. Two basic kinds of check are made for each peak:

  • For each peak any resonances assigned to its dimensions are analysed to highlight potential errors. Some of the checks mirror those found in the “Resonances” tab, e.g. looking for inappropriate covalent bond context. Checks include: missing assignments, like on one side of a covalent pair for an HSQC peak, and seemingly impossible assignments, like having an HSQC peak assigned to two resonances from different residues.
  • The intensity (height and volume) measures for each peak are compared to the average for the whole peak list and significant outliers are highlighted. For magnitude comparisons peak heights and volumes are compared in log-space, and red cells indicate three standard deviations from the sample mean of the logarithm of the intensity. Warnings may indicate the presence of problems, like unintentional diagonal peaks, but some very intense or very weak peaks may be expected. A check is also made if the peak sign seems inappropriate e.g. peaks of opposite sign in a NOESY experiment.

The user may eliminate rows from the table which do not carry any warnings by selecting the “Show only alerts” option.

Caveats & Tips

Because the analyses are relatively slow to perform the tables will not automatically update every time there is an assignment change. If significant changes have been made and the tables need to be updated the user can click the [Update Table] button to force a refresh.

The [Show Peaks] and [Show Resonances] buttons can be used to display all of the peaks and resonances, in tabular form, from which the information in the selected rows are derived. These options are not available for the first tab.

Lots of red warnings in the “Resonances” tab, especially in in the “Bound” column, may be an indication of a problem with the specification of one or more experiment types, rather than a direct problem with the peak assignments. Knowledge of which experimental dimensions represent ‘onebond’ transfers, by virtue of peak assignment, indicates which atomic assignments are in a covalent bond context. The most common mistakes in this regard are the incorrect choice of experiment type or an incorrect choice of dimension order (i.e. how the real experiment dimensions map to the CCPN reference). Both of these problems may be fixed by considering the “Experiment Types” tab in the Experiments popup window. Here, most dimension order problems are fixed by editing the “Transfers To Dim” column in the lower left table, using the user’s knowledge of how the dimensions are connected/correlated.

Main Panel

button Clone: Clone popup window

button Help: Show popup help document

button Close: Close popup

Atom Assign Status

Documentation missing

pulldown Mol System: Selects which molecular system (group of chains) to display assignment status information for

pulldown Shift List: Selects which shift list is used as the source of chemical shift information

entry Residue Selection: Sets which chain and range of residues to show assignment status for, e.g. “2-10, 13-77” or “A1-99, B20-120”

Table 1
Category The name of a category into which atoms of the selected chain/residue have been grouped
Available The total number of atoms (or equivalent sets for methyl groups) that are available for assignment within the category
Assigned The number of atoms (or equivalence groups) that have actually been assigned within the category
% Assigned The percentage of available atoms (or equivalence groups) hat have been assigned within the category
Min Shift The lowest chemical shift value of the assigned atoms within in the category
Mean Shift The average chemical shift value of the assigned atoms within in the category
Max Shift The highest chemical shift value of the assigned atoms within in the category

Through-Space Status

Documentation missing

pulldown Mol System: Selects which molecular system (group of chains) to show through-space connectivity information for

pulldown Peak List: Selects which through-space peak list(s) to show connectivity information for

Table 2
Seq Element The identity of the chain or residue that the data in the row pertains
Total The total number of though-space connections for the chain/residue in the selected peak list(s)
Intra Residue The number of connections that lie within a single residue; do not cross between different residues
Inter Residue The number of connections that cross from one residue to another
Sequential The number of connections between residues that are sequential neighbours
Short Range The number of connections between residues that are within four sequence positions of one another
Short Non-seq The number of connections between residues that are not sequential neighbours but are within four sequence positions of one another
Long Range The number of connections between residues that are five or more sequence positions apart
Intra Chain The number of connections that lie within a single molecular chain
Inter Chain The number of connections that cross between different molecular chains
Contacted Residues For a given residue, a list of sequence numbers for all other residues that are connected

Resonances

Documentation missing

pulldown Shift List: Selects which shift list to derive chemical shift values from in the resonance analysis

check Show only alerts: Sets whether to show only rows in the table for resonances that carry warnings

Table 3
# The serial number of the resonance
Iso. The nuclear isotope type of the resonance
Resonance The assignment annotation for the resonance; just its serial number in square brackets if unassigned
Bound The assignment annotations of all the other resonances that are considered covalently bound; either by virtue of atom assignments or “onebond” linked experimental dimensions
BMRB Mean When atom assigned, the average chemical shift for the atom type given in BioMagResBank data
Random Coil When atom assigned, the sequence-adjusted random coil chemical shift value for the atom type
Type Score When atom assigned, a score indicating how well the resonance chemical shift fits with the known distribution of shifts for the atom type
Value The chemical shift value of the resonance, in the selected shift list
SD The standard deviation of the chemical shift value; over the peak dimensions to which the resonance is assigned
Max Peak Δ The maximum difference in the averaged chemical shift value and a peak dimension assigned to the resonance
Dim Contribs The number of peak dimensions the resonance is assigned to

Peaks

Documentation missing

pulldown Peak List: Selects which peak list to show peak analyses for

check Show only alerts: Sets whether to show only rows in the table for peaks that carry warnings

Table 4
# The serial number of the peak in the selected peak list
Assignment The resonance assignment annotation of the peak
Assign Errors Any error that relate to the peak’s resonance assignment
Location The position of the peak, in the units of the experimental axes; usually ppm
Volume The integration volume of the peak; colour indicates deviation from the peak list average on a logarithmic scale
Log Vol Dev The standard deviation of the peak integration volume on a logarithmic scale; highlights strong/weak outliers (which could be genuine)
Height The height intensity of the peak; colour indicates deviation from the peak list average on a logarithmic scale
Log Height Dev The standard deviation of the height intensity on a logarithmic scale; highlights strong/weak outliers (which could be genuine) and inappropriate peak sign

button Update Table: Manually update the currently viewed table in light of assignment changes etc.

button Show Peaks: Show a table of the peaks that relate to the selected rows, either directly or by virtue of assignment (depending on the current table)

button Show Resonances: Show a table of resonances that relate to the selected rows, either directly or by virtue of assignment (depending on the current table)

Table Of Contents

Previous topic

Chart : Assignment Graph

Next topic

Spin System Type Scores

This Page