Jonas Magnusson, prsenterar sitt examensarbete

Analysis and detection of protein peaks for SELDI-TOF spectra 
  
Abstract:	
	
Advances in the last few years in mass spectrometry have made it
possible to investigate proteins with a wide array of molecular
weights. It's been proposed that this technique makes it possible to
detect dissimilarities in the proteinstructure between cancer
inflicted people and normal people. Since early detection of cancer is
crucial for the treatment this would be a major breakthrough if it
worked.
 
In this thesis we have analysed different mass spectra from SELDI-TOF
technology. Using Bayesian statistics and MCMC methods we developed a
simple model for the spectra. Some parts of the spectra were cut off
due to limitations of the technology, we reproduced these parts to
their assumed real values. Furthermore we developed a method using our
model to compare different groups of spectra. This can be used to
detect protein peaks in the spectra. To reduce more noise from our
data we also, without success, tried an alternative model for our data
based on the assumption that there is a delaying factor within the
equipment.