Detecting Common Insertion Sites in Cancer Research
Author Information
Author(s): Jeroen de Ridder, Anthony Uren, Jaap Kool, Marcel Reinders, Lodewyk Wessels
Primary Institution: Delft University of Technology
Hypothesis
Can a new framework effectively identify common insertion sites (CISs) in retroviral insertional mutagenesis screens while controlling for false detections?
Conclusion
The study introduces a kernel convolution framework that successfully identifies novel common insertion sites in cancer research while controlling for false detections.
Supporting Evidence
- 53% of common insertion sites did not reach the significance threshold in the combined setting.
- The method discovered eight novel common insertion sites with a probability of less than 5% of being false detections.
- The framework allows for analysis at any biologically relevant scale.
- Control of family-wise error rate was maintained throughout the analysis.
Takeaway
Researchers found a way to spot important cancer-related genes by looking at where viruses insert themselves in mouse DNA, even when there are lots of data to sift through.
Methodology
The study used a kernel convolution framework to analyze insertion data from multiple retroviral screens, applying various kernel functions to identify common insertion sites.
Potential Biases
Potential bias due to preferential insertion sites near transcription start sites.
Limitations
The method may not account for all biases in the data, and the background correction model is based on limited information.
Participant Demographics
Mice used in retroviral insertional mutagenesis experiments.
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