Wilhelm

• Ionizer

• Mass Analyzer

• Detector

• Tracing and Labeling to study the movement of substances in systems

• Radiometric Dating

• Medical Diagnostics and Imaging

1. Protein extraction

2. Digestion of protein mixture

3. Peptide seperation

4. MS analysis

5. LC-MS

metric for global confidence assessment of a large-scale proteomics dataset.

FDR = (# Decoy Hits) / (# target hits)

mulitple experiments -> errors accumulate, while correctly identified proteins stay the same

=> After combining, peptide-level and protein-level FDR filtering is necessary to maintain a desired FDR cutoff!

Type 1 - False positive (avoid this one)

Type 2 - False negative

enter m/z values along with their relative abundance

1. Calculate neutral peptide mass

2. Get all peptides which match this mass

3. Generate in-silico spectrum for all peptide candidates

4. Compare and score in-silico generated spectra against experimental spectrum

5. Select best peptide-spectrum-match

Technique that use two or more mass analyzers in sequence => more detailed and specific information about molecules

1. Ionization

2. First Mass Analyzer (MS1)

3. Fragmentation

   collision-induced dissociation (CID), electron capture dissociation (ECD), or infrared multiphoton dissociation (IRMPD)

4. Second Mass Analyzer (MS2)

mass of ion m = m/z value * charge state z

ratio of mass to charge

• isotope spacing

• variation in the numbers of neutrons present in the atom

Molecules with same/similar mass produce a composite peak in low resolution

1. Are two spectra identical

2. Is the top spectrum contained in bottom one?

3. Is the bottom one contained in the top one?

• peak = tupel (m/z & intensity) => spectrum = list of tupels

• m/z of two spectra never the same -> use tolerance to align m/z

  (Is one spectrum contained in another?)

• Normalize: p2-norm - all vectors have the same length of 1

• Similarity measurement: spectran angel, person correlation etc.

for peptides:

use pre-identified spectra -> “move” annotated peaks arount, keep intensity and retention

for metabolite:

not as easy

• Scalability: #proteins increase -> size and complexity increase -> more ressources

• Ambiguity: proteins share peptide sequences -> difficult to assign unique protein ID

• Combinatorial Complexity: complex mix of protein -> many possible combinations of peptides that explain given spectrum

Chromatography based -> LFQ intensity describes area under the peak

Feature: sum of all the MS signals caused by same analyte

Feature Finding: Finding set of features explaining much of signal as possible

Map: 2D-data set (RT, m/z) containing the MS signal from one LC MS run

1. Seeding -> peaks of high intensities

2. Extension -> add peaks arround the seeds

3. Modeling -> estimate parameters of a 2d-feature for the region

4. Refinement -> fit model to collected peaks, remove unmatching ones

5. Iterate until convergence

• Local Maxima/Minima -> peaks and bottoms

• Fitting methods -> try to fit expected shape

• convolution

• 
  

• retention time of the same analyte will fluctuate across machines (slightly

  different LCs) and over time

• Use 2 analytes , one eluting rather early, the other rather late in the gradient

  (LC), in praxis a known peptide is used

• measure similarity of two time series

• calculate distance between two points (e.g. euclidean distance)

• 
  

Identifying pairs of genes in two different genomes which are more similar to each other than either is to any other

-> used for matching peaks

• can be used with dynamic time warping

  
  

RetentionTime will differ

• mass analyser and detector in one

• high resolution at low m/z

• low resolution at high m/z