LS-NMF: A New Algorithm for Analyzing Microarray Data
Author Information
Author(s): Wang Guoli, Kossenkov Andrew V, Ochs Michael F
Primary Institution: Fox Chase Cancer Center
Hypothesis
Incorporating uncertainty measurements into the non-negative matrix factorization (NMF) algorithm will improve its performance in analyzing microarray data.
Conclusion
The LS-NMF algorithm significantly enhances the ability to recover biological information from microarray data by utilizing uncertainty measurements.
Supporting Evidence
- LS-NMF outperformed the original NMF algorithm in identifying functionally related genes.
- The incorporation of uncertainty information improved the recovery of biological knowledge.
- LS-NMF maintained the advantages of the original NMF while enhancing performance.
Takeaway
Scientists created a new tool called LS-NMF that helps find patterns in gene data better by using extra information about how uncertain those data points are.
Methodology
The LS-NMF algorithm integrates uncertainty measurements into the NMF updating rules to improve the identification of functionally related genes.
Limitations
The algorithm may not perform optimally in cases where no variance information is available, as it reduces back to the original NMF.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website