An automated rule based approach to MS/MS spectra interpretation: Discovery of novel mitochondrial inner membrane proteins

Ioannis Xenarios
Serono International

Over the past decades mass spectrometry has generated thousands of mass spectra. It is no surprise that in the pharmaceutical and biotech industry mass spectrometry plays a key role in the quality control of compounds (being proteins or small molecules). It is now also a fact that the industry has moved towards characterizing biological fluids, cellular compartments, and biochemical complexes using high throughput mass spectrometry. However the major hurdle generated by this avalanche of data is the correct interpretation of MS or MS/MS spectra. We tackled the problem of data interpretation using our mass spectrometrist's expertise to annotate a set of MS/MS spectra and to categorize and define the rules and 'tricks' used to interpret these data. Together this information allowed the computational biologists to develop a tool that mimics the human know-how. We present an example of a high throughput experiment where preparations of the inner membrane of the mitochondria were submitted to digestion and subsequently analyzed by mass spectrometry. Using our approach allowed us to interpret higher fraction of MS/MS spectra with a higher accuracy than industry standard and to discover novel mitochondrial inner membrane proteins.