Biclustering Gene Expression Data Using Non-Smooth Non-Negative Matrix Factorization
Author Information
Author(s): Carmona-Saez Pedro, Pascual-Marqui Roberto D, Tirado F, Carazo Jose M, Pascual-Montano Alberto
Primary Institution: National Center of Biotechnology, Spain
Hypothesis
Can the non-smooth non-negative matrix factorization (nsNMF) method effectively uncover local structures in gene expression data?
Conclusion
The nsNMF method is effective in identifying complex relationships among genes and conditions in large gene expression datasets.
Supporting Evidence
- The nsNMF method successfully identified localized features in synthetic datasets.
- The method revealed meaningful biological insights from large gene expression datasets.
- Functional annotations were significantly over-represented in identified gene modules.
- The approach outperformed standard clustering algorithms in detecting complex patterns.
Takeaway
This study shows a new way to group genes that work together in specific situations, helping scientists understand how genes behave in different conditions.
Methodology
The study used non-smooth non-negative matrix factorization (nsNMF) to analyze synthetic and real gene expression datasets.
Limitations
The method may not capture all relationships in datasets with very complex structures.
Participant Demographics
The study analyzed gene expression data from various human tissues and soft-tissue tumors.
Statistical Information
P-Value
0.0
Statistical Significance
p<0.01
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website