Results of matrix factorization decompositions for a set of 14 colon cancer transcriptomic datasets using different methods

These are the results of application of several matrix factorization methods (ICA, StabilizedICA, NMF, PCA) for a set of 14 independent colon cancer transcriptomics datasets. The analysis was done to compare the methods in terms of their reproducibility (ability to generalize). 

The results of this analysis was published in 

Cantini L, Kairov U, de Reyniès A, Barillot E, Radvanyi F, Zinovyev A. Assessing reproducibility of matrix factorization methods in independent transcriptomes. Bioinformatics. 2019 Nov 1;35(21):4307-4313. doi: 10.1093/bioinformatics/btz225.