Skip to main content
Fig. 1 | BMC Bioinformatics

Fig. 1

From: Systematic noise degrades gene co-expression signals but can be corrected

Fig. 1

Density curves of the correlation between random genes for differently cleaned versions of all datasets. a Correlation density curves as estimated from the raw, BC, BC+QN and RUV-random versions of the Hawrylycz et al. data. The dashed line (purple) represents the density of the correlations between random genes estimated from the version of the data normalized by the authors. b Correlation density curves estimated from the raw, BC, BC+QN and RUV-random versions of the Miller et al. data. The dashed line (purple) represents the density of the correlations between random genes estimated from the version of the data normalized by the authors. c Correlation density curves estimated from the raw, BC, BC+QN and RUV-random versions versions of the Kang et al. data. d Correlation density curves estimated from the raw, BC, BC+LN and RUV-randomversions of the Colantuoni et al. data. e Correlation density curves estimated from the raw, BC, BC+QN and RUV-random versions of the Hernandez et al. data. The dashed line (purple) represents the density of the correlations between random genes estimated from the version of the data normalized by the authors. f Correlation density curves estimated from the raw, BC, BC+QN and BC+RUV-random versions of the Hernandez et al. data. Note that here we display BC+RUV-random instead of RUV-random. The histogram in the background of all panels represents the density generated when using the raw version of each dataset

Back to article page