Related Experiment Videos
Is there a bias in proteome research?
1Johannes-Müller-Institut für Physiologie, Humboldt-Universität zu Berlin, Berlin, Germany. mrowka@rz.hu-berlin.de
Genome Research
|December 4, 2001
Summary
Comparing proteomics data reveals significant differences between individual studies and genomewide scans. Protein interactions identified in single publications show closer transcriptional profiles than those from large-scale genomewide analyses.
Area of Science:
- Proteomics
- Systems Biology
- Bioinformatics
Background:
- Technological advancements allow re-evaluation of traditional experimental data alongside genomewide datasets.
- Significant discrepancies exist between data from single experiments and large-scale genomewide studies in proteomics.
Purpose of the Study:
- To compare protein-protein interaction datasets in Saccharomyces cerevisiae obtained through identical technical methods.
- To investigate the differences between individually identified interactions and those from genomewide scans.
Main Methods:
- Utilized the yeast two-hybrid system for detecting protein-protein interactions.
- Compared data sets from single publications with two genomewide scans of protein interactions in Saccharomyces cerevisiae.
Main Results:
- Individually identified protein-protein interactions differ substantially from those identified by genomewide scans.
- Interacting proteins from pooled single publications exhibit closer transcriptional profiles compared to genomewide data.
- Analysis suggests potential for high false positive rates (44-91%) in genomewide scans or significant selection bias in single publications.
Conclusions:
- Genomewide scans in proteomics may contain a substantial fraction of false positives.
- Alternatively, hypothesis-driven experiments might involve a significant selection process, influencing identified protein interactions.
- Reconciling differences between single-study and genomewide proteomics data is crucial for accurate biological interpretation.