Plotting script — display the principal components identified by the main script

• GUI and CLI programs

  GUI: trendplotgui.exe in Windows and Linux, trendplotgui.app in OS X

  CLI: trendplot.exe in all platforms

• Synopsis

  trendplot.exe [-r] [-f] [prefix] [-x [xaxisfile]] [-u [Unit]] [-n [the first N components]] [-p] [-e] -m [0/1/2/3/4/5]] --plot

• Description

  • trendplot.exe plots the output of trendmain.exe. It is normally used to plot first principal component (PC1) or n principal components. It can also graph a scree plot of the sizes of the PCs. It uses matplotlib.
  • Note that each time trendmaingui or trendmain launches it creates a temp directory to save temporary files for trendplot and trendplotgui and deletes the old one (if it exists). This enables trendplot and trendplotgui to use results of the last run of trendmain or trendmaingui.

• Options

  • -r
-r reads all arguments used by trendmain.exe. Therefore when -r is used all trendplot.exe options related to trendmain.exe program, such as -f and --ica options can be omitted.
  • -f [input prefix name]
-f specifies prefix of input files. This needs to be the same as specified with the -o option when using the trendmain.exe program. e.g. uM or time (min)
  • -x [xaxisfile]
    The usage is identical to the -x option of trendmain.exe.
  • -u [Unit]
    The usage is identical to the -u option of trendmain.exe.
  • -n [integral number N]
-n specifies the number of principal components to plot. For example, -n 3 plots the first 3 principal components.
  • --single
--single plots the component number N specified by -n option. For example, -n 3 --single plots the third principal component only.
  • -p
-p plots PC1 values with normalization. This is normally the most important trend of change in an experiment. It is used without argument.
  • -e
-e plots the scree plot, i.e. the size of each of the principal components
  • -m [0/1/2/3/4/5]-m turns on normalization of the N principle components plotted using the -n option. There are several modes of normalization to choose among; see the table below for the choices. When -m option is turned on, the normalized PC values will be saved to normalizedPC.csv file.

Normalization mode	meaning
0	No normalization on PC values
1	Normalize all PC values to 1.0, with PC1 increasing (typical for a ligand binding isotherm)
2	Normalize all PC values to 1.0, regardless of the direction
3	Normalize PC1 to 1.0 and set it to increase. Normalize all other components to PC1
4	Normalize PC1 to 1.0 without changing its orientation. Normalize all other components to PC1
5*	Normalize PC values to the sum of PC1+PC2
6$	Normalize 2 phase binding isotherms into monotonic increasing, reverse turning point

* This mode was used for two sequential binding events
$ This option might be used for binding isotherms of titrations where spectra in the middle of the titration are severely perturbed and the trajectory of change is V-shaped.

• --biplotThis option shows the “bi-plot” of the first two principal components in matrix V^T. (A bi-plot is the equivalent of a score plot in the nonmenclature of the metabolomics field).
• --icaThis option is required for plotting ICA results. All principal components mentioned above will be replaced by independent components when --ica is turned on.
• --plot
--plot is optional. When it is on, all figures will be saved as PNG images.