New Clustering Method for Gene Expression Data
Author Information
Author(s): Pankaj Chopra, Jaewoo Kang, Jiong Yang, HyungJun Cho, Heenam Stanley Kim, Min-Goo Lee
Primary Institution: North Carolina State University
Hypothesis
Can a new clustering model generate multiple versions of different clusters from a single dataset to reveal new biological insights?
Conclusion
The SigCalc algorithm reveals new, biologically meaningful clusters that were not evident in conventional clustering methods.
Supporting Evidence
- The SigCalc algorithm can generate different clusters from the same dataset using different landmark genes.
- New biological associations were revealed that were not apparent from conventional clustering methods.
- The study demonstrated that many unique GO terms were common across different datasets.
Takeaway
Scientists can use a new method to look at gene data in different ways, helping them find new connections between genes.
Methodology
The study used the SigCalc algorithm to project microarray data onto a subspace defined by landmark genes and applied clustering techniques to identify gene associations.
Limitations
The study is limited to the datasets used and may not generalize to other biological contexts.
Statistical Information
P-Value
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website