If you have calibration data for a particular reference compound, you must create a csv file and folder that have the same base name as the reference MS file from above. Again, an examples are provided in this repository called refcpd.csv and the folder refcpd. All of your calibration AIA files for this compound need to be stored in the newly created folder. In order for these new data files to be processed, the refcpd.csv file must be appropriately modified.
The csv file is a simple comma-separated text file, but again the structure is important. The first row in this file is critical. At the end, there are two values that define the starting and stopping time points for integration. Change these values based on the time range that you’ve determined from the TIC of a calibration run. The rest of the rows are data file information. The first column is the name of a calibration data file, and the second column needs to be the concentration of the reference compound associated with that run. You don’t have to add all of the calibration files here, but if they are not in this list, they won’t be processed. Alternatively, any line that starts with a ‘#’ is a comment, and will be ignored. In this way, you can comment out samples, and add some notes as to why that sample was not used or whatever.
Once you’ve updated the calibration information from above. You can run the program ‘calibration.py’. This runs through all of the reference spectra defined in the ‘reference_files.txt’ file. If a ‘.csv’ file exists for a particular reference file, then a calibration will be performed.
All of the calibration data files listed in the csv file will be processed and a calibration curve generated. For each calibration sample, a plot of the reference-extracted data will be generated in the calibration folder (refcpd_fits.png). In addition, a calibration curve plot is also generated (refcpd_cal_curve.png’), which plots the integrated intensities and calibrated intensities vs the concentrations. In addition, the calibration information is printed on the graph for quick visual inspection. There is no need to write down this calibration information.
This program has some important command line arguments that will change the programs defaults. The first argument, ‘–nobkg’, is a simple flag for background fitting. By default, the fitting routine will select a MS slice from the data set and use that as a background in the non-negative least squares fitting. This procedure can change the integrated values. If you use this flag, then a background MS will not be used in the fitting. Using a background slice in the fitting may or may not give good results. It might be a good idea to look at your data with and without the background subtraction to see which is better.
The second command line argument is ‘–bkg_time’. By default, the fitting program uses the first MS slice as a background for fitting. However, if there is another time that looks like it might make a better background for subtraction, then you can put that number here.
Here’s a couple of example usages of this script:
# This will run the calibration program with all defaults
$ python calibration.py
# This shuts off the background subtraction
$ python calibration.py --nobkg
# This sets an alternate time for the background subtraction
# In this case, the time is set to 0.12 minutes
$ python calibration.py --bkg_time 0.12
Another file is also generated during this process: cal.h5. This is a HDF5 file that contains all of the calibration information for each standard. Do not delete this file; it is essential for the next step. This is a very simple file, and there are many tools for looking at the internals of an HDF5 file. For example, ViTables is recommended. The background information, such as whether a background was used and the time point to use as a background spectrum, are stored as user attributes of the calibration table.