Heat map and visualization settings

After the “Accept” button had been pressed, a tab for each lipid class appears. This tab again contains a tabbed pane with the entries “Heatmap” and “Bar-chart”, and if groups are selected “Group-Heatmap” and “Group bar-chart”. The “Heatmap” tab is selected by default. The tab itself contains the heat map at the top, followed by an export bar, and some control elements at the bottom.

Heat map:

At the top, the heat map contains a color legend. The values in the map are calculated relative to the median of one molecule over all samples/groups. If the intensity of a molecule of one sample/group is lower than the median, it is colored green, the ones that are higher are red, and the ones around the median are black. If one analyte could not be quantified it is gray. In the heat map the samples/groups are organized horizontally and the lipids vertically. If a quantitation is based on more than one peak (could be an ambiguous identification; each isotope separately), a yellow rectangle is around the peak. If the mouse is hovered over one heat map cell, a white rectangle is rendered and at the bottom of the application a line with the name of the lipid, the name of the sample, the value relative to the median, a value depending on the “Settings” option (can be standardized), the original value of the quantitation, and the amount of available isotopes is displayed.

Mouse clicks on cells in the heat map (not valid for the group heat map):

- Left mouse button (does not work for gray fields): The LDA visualizes the quantitation (see Display Results). If the quantitation is based on several peaks due to adducts/modifications, the LDA jumps to the first one it finds.
 
- Right mouse button (does not work for gray fields an the ones with a yellow rectangle): A popup menu appears with 2 options: 1) “Choose just one peak for doubles”; this option is for the automated removal of double peak identifications (yellow rectangle). It automatically selects in the other samples the one peak as correct, that is closer to the retention time of the sample where the right mouse button has been clicked. After this option has been selected the application asks for which adducts/modifications this operation should performed.

2) “Quant. anal. at not found”; this option tries to automatically quantify analytes which are grey at the retention time of the selected analyte, whereby it does not take the theoretical isotopic distribution into account. This procedure tries out all of the available quantitation methods, starting with the 3D method, then the “greedy” method followed by the MASPECTRAS and ASAPRatio method (see chapter Chromatogram viewer). For this option the application asks as well for which adducts/modifications this operation should be performed like in “Choose just one peak for doubles”.

Mouse clicks on the “sample/group” name:

- Left mouse button: a bar chart for the sample/group is displayed, containing all the available analytes.
 
- Right mouse button (just for sample): Renaming of sample name.

Mouse clicks on the molecule name:

- Left mouse button: a bar chart for the molecule is displayed, containing all available samples/groups.
 
- Right mouse button (just valid for sample): A popup menu appears containing 3 options: 1) “Remove analyte in all samples”: deletes one lipid in all result files. 2) “Select analyte”: If there are a lot of lipids and a lot of result files, the removal of a lipid can be quite time consuming. However, if there are several analytes deleted, it consumes the same time. With this option analytes can be selected for the removal; if selected, the name of the analyte is surrounded by a blue rectangle. The removal takes effect on all of the selected analytes as soon as the “Choose just one peak for doubles” option is pressed. Here again a confirmation box appears asking for the adducts/modification on which the removal should be performed.

Export bar:

The picture of the heat map or the data respectively can be exported in the following formats: PNG = portable network graphics; SVG = scalable vector graphics; Excel = Microsoft Excel; Text = text based format (tab delimited). Excel and Text export the values in the type that are selected by the user (see next paragraph - Control elements). Furthermore, there is the option to export results in mzTab format, in order to submit them to a public repository. For mzTab, the raw peak areas are exported, these can be affected by standardization and isotope setting only. The mzTab button exports all lipid classes at once and generates a file containing all experiments; for the manual validation, the data can be exported to chromatograms with the "Chroms" button. Here, the exported analytes, experiments and modifications can be selected.

It is recommended to export just one or two lipid series at once. The chroms export can take some time, depending on the data and the amount of selected hits. The progress of the export is presented in a progress bar underneath the heat map.

At the end of the chroms export, the user is informed that the export is finished.

This picture is a chroms export of TG52 of the LDA. It can be easily seen that analytes with more double bound elute slightly earlier. This is a good quality criterion that the algorithm selected the correct hit.

Control elements:

- show intern. stand.: should the internal standards be displayed in the heat map
 
- show extern. stand.: should the external standards be displayed in the heat map
 
- isotopes: what is the highest isotope number that should be used for the heat map for all of the analytes. However, the highest isotope number is determined for each analyte separately (e.g. if we have three samples 1, 2 and 3 and for 1 and 2 the isotopes +0,+1, and+2 are found, whereas in sample 3 just +0 and +1 are found; just +0 and +1 are taken for the heat map).
 
- double peaks: should duplicate peak identifications be flagged with the yellow rectangle
 
- Settings button: This is the major settings panel for the display.

      > value type: which value should be displayed and used for the heat map or the bar chart.
            o   relative value: Just a value in arbitrary units. This value could be the calculated quantity for the analyte itself, or it could be standardized on standards, depending on the other settings.
            o   relative to base peak: The values are calculated relative to the highest found peak of this lipid class.
            o   relative to measured class amount: Here percentual values are determined relative to the sum of all analytes for one sample (standards are not considered in this sum) of one lipid class.
            o   relative to highest total peak: The values are calculated relative to the highest found peak of all available lipid classes.
            o   relative to total amount: The values are calculated relative to the highest found peak of all available lipid classes.
            o   amount end volume: Returns the amount [mol] in the end volume, before the MS measurement. For this value standardization on an internal standard and absolute quantities (see Results selection and settings number 5) are required.
            o   conc. end volume: Returns the concentration [mol/L] in the end volume, before the MS measurement. For this value standardization on an internal standard and absolute quantities (see Results selection and settings number 5) are required.
            o   weight end-volume: returns the weight [gram] in the end volume, before the MS measurement. For this value standardization on an internal standard and absolute quantities (see Results selection and settings number 5) are required.
            o   amount sample-volume: Returns the amount [mol] in the sample volume, before the sample preparation steps (dilution). For this value standardization on an internal and/or external standard and absolute quantities (see Results selection and settings number 5) are required.
            o   conc. sample-volume: Returns the concentration [mol/L] in the sample volume, before the sample preparation steps (dilution). For this value standardization on an internal and/or external standard and absolute quantities (see Results selection and settings number 5) are required.
            o   weight sample-volume: Returns the amount [gram] in the sample volume, before the sample preparation steps (dilution). For this value standardization on an internal and/or external standard and absolute quantities (see Results selection and settings number 5) are required.
            o   relation to measured neutral lipid: This is for the standardization on the total measured lipid mass[g] of the current lipid class measured by MS. For this value standardization on an internal and/or external standard and absolute quantities (see Results selection and settings number 5) are required.
            o   relative to sample weight: This is for the standardization on the total weight of the sample mass[g] of the current lipid class measured by MS. For this value standardization on an internal and/or external standard and absolute quantities (see Results selection and settings number 5) are required.
            o   relation to protein content: This is for the standardization on a differently measured protein mass (e.g. measured by a kit). The values for the standardization have to be entered at the absolute quantities (see chapter Results selection and settings number 5). If there is a standard available, this value is presented in mol/g, else in AU/g.
            o   relation to neutral lipid content: This is for the standardization on a differently measured lipid mass (e.g. measured by a kit). The values for the standardization have to be entered at the absolute quantities (see chapter Results selection and settings number 5). If there is a standard available, this value is presented in mol/g, else in AU/g.

      > internal standard correction: is the standardization on an internal standard desired. If yes, it is possible to choose between 3 standardization methods.
            o   most reliable standard: Internally developed method, that detects trustworthy standards and out of these a ratio to the other groups is calculated.
            o   median: The median of the standards of one sample is taken as reference value for standardization
            o   single standards (several ones are available e.g. IS50:0): A standardization on every entered standard is possible.
      > consider dilution: Consider the dilution in the calculation.
      > divisor unit: If the value is standardized by some value, this value defines the magnitude of the divisor (e.g. pmol/mL; the m for milli is defined by this setting). This setting is applicable for "conc. end-volume", "conc. sample volume", "relation to measured neutral lipid", "relative to sample weight", "relation to protein content", and "relation to neutral lipid content".
      > use AU: use arbitrary units instead of SI units.
 
- Select molecules: To fade out specific molecules of the heat map.
 
- Combined chart: Shows selected molecules in one chart over all samples/groups (see Bar chart).
 
- Export options: Defines the information that is exported to Excel or text format. This dialog box consists of a radio button with the options “analytes in column” and “experiments in column”, and additionally a checkbox for the export of the retention time. Additionally for groups, the standard deviation or the standard error of the values and the standard deviation of the retention time can be exported.