Effects of Incomplete Protein Interaction Data
Author Information
Author(s): Eric de Silva, Thomas Thorne, Piers Ingram, Ino Agrafioti, Jonathan Swire, Carsten Wiuf, Michael PH Stumpf
Primary Institution: Imperial College London
Hypothesis
How does incomplete protein interaction data affect structural and evolutionary inferences?
Conclusion
Ignoring the incompleteness of protein interaction data can lead to significant biases in biological analyses.
Supporting Evidence
- The study shows that bias is virtually inevitable when only small, partial network data sets are considered.
- Previous analyses of protein interaction networks may need to be reassessed due to the effects of incomplete data.
- The research highlights the importance of considering network sampling properties in biological analyses.
Takeaway
When scientists study proteins, they often miss some connections, which can lead to wrong conclusions about how proteins work together.
Methodology
The study analyzed the effects of random and non-random sampling schemes on the yeast protein interaction network using various network statistics.
Potential Biases
Bias may arise from the incomplete nature of network data, affecting the understanding of biological systems.
Limitations
The study primarily focuses on qualitative analysis rather than quantitative assessments of the reliability of the dataset.
Participant Demographics
The study focuses on the protein interaction network of Saccharomyces cerevisiae.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website