Swiss Institute of Bioinformatics

Saving Patients using Mass Spectrometry (SPUMS)

---

Background

The negative effects of drugs that are supposedly correctly prescribed is a major cause of death in many western countries. Technically, this situation is described as the worse case of Adverse Drug Reaction (ADR). These effects are suspected to arise from unexpected interactions between various drugs and unknown effects due to phenotypic and genetic predisposition in patients. A drug target is generally a protein playing a key role in a particular biological process. In ADR, chemicals are likely to affect target proteins at the molecular level through subtle chemical changes causing toxicity. Therefore, the detection of drugs in body fluids should ideally be completed by the detection of potentially affected proteins.

SPUMS project

Mass spectrometry (MassSpec) is an analytical technique for the determination of the elemental composition of a sample. While this technique is increasingly introduced in clinical settings, no large-scale screening method for assessing the effect of ingested drugs on circulating proteins is currently available. Data management, software processing and data visualisation are significant limiting factors. Our goal is to develop a bioinformatics platform that can automatically detect the effect(s) of the variable presence of drug chemical fragments obtained from patients? blood samples on circulating peptides in numerous mass spectra.

Major experimental limitations

1. Identification of post-translational modifications (PTMs) is complex
2. Dynamic range of protein concentration (>10¹⁰): need for depletion, i.e. loss of abundant proteins

The main computational challenges belong to 3 categories: (i) algorithms for large dataset analysis (ii) management, modelling and data storage (iii) accession and visualisation of data. Data is managed with SQLServer. The database transforms clinical, MassSpec and identification data from standard input formats into an efficient storage system. SIB and MSR are closely collaborating to meet the project complex data management and visualisation needs. Data (peak lists, identified peptides, etc.) will also be explored and processed to assist scientists in investigating drug effects on target proteins via further analysis.

Summary and details of deliverables:

1. Clinical protocol to produce spectra from blood samples.
2. Relational database of detectable peptides in associated mass spectra related to functional characteristics of corresponding proteins.
3. Matching tools for characteristic MassSpec profiles of patients with reference MS.
4. Performance evaluation by statistical tests and comparison with control data.
5. Visualisation tools and data/knowledge transfer processes between similar distributed platforms.

The project, if successful, should result in a significant impact in hospitals worldwide.

Last modified 20/Mar/2009 by PAP