By Phil Holt of the Monash University LC/MS Unit
The SpectraPhile ESI MS Deconvolution applet
If you can't see the applet, then you need Java, get it at www.java.com
- What is?
- [Threshold]
- [Tolerance]
- [Min. output mass]
- [Deconvolution]
- [How to use]
- [How it works]
- [Limitations]
- [Download and save]
- [About SpectraPhile]
The applet allows you play with a sample ESI spectrum of a snake venom protein, or paste in your own text spectral data and view it and a deconvoluted spectra. It also gives you the opportunity to play with some of the input parameters. The applet is limited to cut and paste due to the security restrictions of your internet browser preventing it from opening files or directly accessing the clipboard on your system.
Update version 1.2
As of June 2006, the applet now formats the numeric text in the output window to limit decimal points and construct two orderly columns.
How To Use Applet
- First convert your spectrum to centroid data to minimise file size
- Mass Spectroscopy software like MassLynx enables you to 'copy a spectrum list.' Use this facility in your MS software to copy a spectrum to the clipboard as a list then paste into the browsers input text area below. Your text must appear as 2 columns of data (mass magnitude pairs).
- Press Start
- Choose some parameters form the drop down combo boxes
- Press Deconvolute
- View output spectrum. Try different parameters and press deconvolute again if necessary.
- Select and copy output text and use a text editor to paste and save it to a text file.
- To enter more data, first clear the text areas.
Deconvolution algorithm, how it works
It is similar to manual mass deconvolution. Every peak over a set threshold is compared to every other peak. For each pair, divide the difference into the higher mass peak to obtain the charge on the lower. If the charge close to a whole number (within the chosen tolerance, then they both come from the same molecular ion. If the calculated charge is not near a whole number then ignore it. The whole number charge is used to multiply the value of lower mass peak to obtain the molecular mass. More matches in the same series increase the number of peaks clustered around the molecular mass peak.
Limitations
The program only considers proton adducts, ie. H+. No attempt is made to calculate peaks using Na+ or NH4+ or K+.
Artifact peaks occur at 1/2 the molecular weight in many cases. This is due to what should be adjacent charge envelope peaks failing the tolerance test, but the next ion in the series passes. So the program calculates on the basis of the wrong charge. By adjusting to a higher value tolerance you can minimise this effect.
Threshold
The ratio of the examined peak height over the largest peak height, expressed as a percentage. Set it to 10 or 25% to ignore all of the small peaks.
Tolerance
The tolerance value is the margin between the whole number and the calculated charge that is acceptable. eg the difference between 5.995 and 6 is 0.005. If tolerance is set to 0.001 this match would be ignored.
Min output mass
This is the minimum mass value for the deconvolution output. If you set it too high you may miss some peaks.
Back to topDownload and save applet
If you download the java applet, save it on your machine, then run it as an application, you will be able to run SpectraPhile as an application on your machine, open text files of spectra, deconvolute and then save files directly on your disk.
Saving instructions
Click on SpStart.jar.Save the file in a directory called SpectraPhile
Operation instructions
Depending on your operating system-
In Windows, you can locate the SpStart.jar file via say Windows Explorer, then double click on it. (Jar files are self extracting providing you have the right Java Runtime Environment Version).
Alternatively, call up the command line prompt (eg MSDOS)
change to the SpectraPhile directory
Type after the C:\SpectraPhile>java -jar SpStart.jar
Hit enter, and wait several seconds for the application GUI to load.
About the SpectraPhile Project
This was started as a practical exercise in my Java subject at Chisholm TAFE. The Classes were slowly built up from a basic deconvolution of typed data trials, through GUI, applet development, then file handling.
It is a simple program and does have a few flaws. Artifact peaks can be generated, so you need to manually check these to see if they have a whole number multiplier relationship with other generated peaks.
I will soon provide a feedback form for your comments, but for now, just send them to:
phillip\dot\holt\at\sci\dot\monash\dot\edu\dot\au
This service was provided by armball.net
Back to top