Page QuantProtP4

This is the project page of the quantitative proteomics group.

Students

Mail an alle Gruppenmitglieder: AA2010SS-QuantProt bei lists.spline.de

Name	email
Christian Hoppe	christian.hoppe@fu-berlin.de
Kersten Döring	kdoering@zedat.fu-berlin.de
Stefan Mendt	mendtste@zedat.fu-berlin.de
Christoph Standfuß	christoph.standfuss@inf.fu-berlin.de
Peter Moor	moor@inf.fu-berlin.de
Maria Kruegle	mariadg@zedat.fu-berlin.de
Anna Kosenko	annakriv@zedat.fu-berlin.de
Matthias Kluge	mkluge@inf.fu-berlin.de
Felix Mattes	mattesf@inf.fu-berlin.de

Literature

Initial reading (this should be read by all)

In this paper you find an overview of label-free quantitative approaches. In addition to label free approaches there are numerous labeling techniques.

Apart from quantitation, peptide and protein ID are another important topic, which we will discuss. Most database search engines are based on the spectrum graph, which was initially introduced by Dancik et al in this paper.

For the core lecture you should read the Dancik paper.

Further reading

Here you find material that is connected in the quantitative analysis of proteomics data.

• Labeling techniques
• Peptide/Protein ID
  • Peptide identification, Inspect
  • Peptide identification, Scope
  • Dynamic Programming for spectrum graph
• Signal processing/Peak picking
  • Isotope distribution and mass decomposition
  • Signal processing for proteomics
  • Peak Picking wavelet

Exercises

Exercise on peak detection

The project

Proposal outline

• Signal processing
  • Overview - low-level preprocessing
  • Precision, accuracy and resolution
  • Peak picking
  • Isotopic patterns

• Peptide feature detection
  • Alignment of LC-MS data
  • database approaches

• Peptid ID
  • de novo sequencing
  • spectrum graph

Analysis/Programming Projects

Possible projects

Please note, that if you are interested in one of those we still have to work out the details. These are some ideas.

1. Comparing elution profiles [Bielow](Details) (Stefan, Peter) Elution profiles of features are usually roughly gaussian shaped and correlate if they stem from the same analyte. This can happen due to multiple charge variants as observed in ESI or adduction of salt ions. The goal of this project is to compare elution profiles based a correlation measure, given a list of corresponding features.
2. emPAI Status: finished [Andreotti](Details) (Anna) The "exponentially modified protein abundance index" can be used to quantify proteins based on the peptides that were observed. This also involves prediction of "proteotypic" peptides (the ones that ionize well) using machine learning techniques. (http://www.ncbi.nlm.nih.gov/pubmed/15958392)
3. Simulation of PTM's Status: finished [Bielow](Christian,Kersten) (Details) AMS 3.0 is a software using artificial neural networks and voting to predict modification sites. Predicting PTMs in simulated experiments would add another dimension of realism to the results. (http://www.biomedcentral.com/1471-2105/11/210/abstract)
4. Mass fingerprinting [Bielow](Details)(Christoph) :To get a feeling for how "unique" a certain peptide mass is, try to digest (in-silico) a proteome (e.g. from human) using trypsin. How many peptides are unique to a single protein? Given a mass precision of x ppm, how many peptides are uniquely identifyable? What happens we additionally allow 1 or 2 PTMs? …
5. Baseline estimation [Bielow](Matthias) One of the crucial steps in MALDI data processing, an algorithm by Williams et al seems promising (http://portal.acm.org/citation.cfm?id=1167394)

• Proteomics_2008_AmericaComparative_LC-MS_A_landscape_of.pdf: Comparative LC-MS
• J_Comput_Biol_1999_DancikDe_novo_peptide_sequencing_via.pdf: Spectrum graph
• proteomics-ms-overview.pdf: Proteomics Overview
• isotope-distribution.pdf: Isotope distribution and mass decomposition
• J_Comput_Biol_2003_LuA_suboptimal_algorithm_for_de.pdf: Dynamic Programming for spectrum graph
• peptide-id-inspect.pdf: Peptide identification, Inspect
• peptide-id-scope.pdf: Peptide identification, Scope
• signal-processing.pdf: Signal processing for proteomics
• Pacific_Symposium_on_Biocomputing_Pacific_Symposium_on_Biocomputing_2006_LangeHigh-accuracy_peak_picking_of_proteomics.pdf: Peak Picking wavelet

• Chen_et_al._2001_dynamic_programming_approach_to_de_novo_peptide_sequencing_via_tandem_mass_spectrometry.pdf: Dynamic programming algorithm for spectrum graph
• J_Comput_Biol_2003_LuA_suboptimal_algorithm_for_de.pdf: Extension of Chen's dynamic program for suboptimal solutions
• Fischer_et_al._2005_NovoHMM_a_hidden_Markov_model_for_de_novo_peptide_sequencing.pdf: peptide identification with hidden markov model

• Tanner_et_al._2005_InsPecT_identification_of_posttranslationally_modified_peptides_from_tandem_mass_spectra.pdf: Peptide ID INSPECT paper