Stability-driven nonnegative matrix factorization to interpret spatial gene expression and build local gene networks

Authors:

Abstract
Significance: Despite the abundance of spatial gene expression data, extracting meaningful information to reveal how genes interact remains a challenge. We developed staNMF, a method that combines a powerful unsupervised learning algorithm, nonnegative matrix factorization (NMF), with a new stability criterion that selects the size of the dictionary or the set of principal patterns (PP). We demonstrate that PP give rise to a novel and concise representation of theDrosophilaembryonic spatial expression patterns and they correspond to biologically meaningful regions of theDrosophilaembryo. Furthermore, this new representation was used to automatically predict manual annotations, categorize gene expression patterns, and reconstruct the local gap gene network with high accuracy.
Funding Information
  • HHS | National Institutes of Health (R01 GM076655)
  • HHS | National Institutes of Health (R01 GM097231)
  • National Science Foundation (DMS-1107000)
  • National Science Foundation (CDS&E-MSS 1228246)
  • National Science Foundation (DMS-1160319 (FRG))
  • HHS | NIH | National Human Genome Research Institute (1U01HG007031-01)
  • DOD | Air Force Office of Scientific Research (FA9550-14-1-0016)
  • DOD | Army Research Office (W911NF-11-1- 0114)
  • National Science Foundation (CCF-0939370)

This publication has 47 references indexed in Scilit: