Table 1.
Final merged dataset and training and validation sets.
Fig 1.
Venn diagram showing the overlap and discrepancies between genes differentially expressed in Clarkson et al. [3] and Stein et al. [4].
Fig 2.
The log2 fold change in gene expression profiles for the ten significant clusters identified through the STEM Clustering.
Y-axis represents the relative gene expression levels of involution days 0.5, 1, 2, 3 and 4 days compared lactation day 10 in log base 2 scale. X-axis represents the time points (L10, lactation day 10; I12h, involution day 0.5; I24h, involution day 1; I48h, involution day 2; I72h, involution day 3; I96h, involution day 4). SOCS3, suppressor of cytokine signaling 3; IGF1, insulin-like growth factor 1 (somatomedin C); STAT3, signal transducer and activator of transcription 3 (acute-phase response factor); TGFB3, transforming growth factor, beta 3; ATF4, activating transcription factor 4; IGFBP5, insulin-like growth factor binding protein 5; MMP3, matrix metallopeptidase 3.
Table 2.
Patterns, size and p-values of significant clusters identified through the STEM algorithm using data from Clarkson et al. [3].
The far right column indicates Involution Specific Gene Signatures through the STEM clustering using data from Clarkson et al. [3].
Fig 3.
Enrichment plots form GSEA for Inv5 signature for IBC versus non-IBC in the training and validation sets.
nIBC, non-Inflammatory breast cancer.
Table 3.
GSEA results of involution-specific signatures in IBC versus non-IBC in the training and validation sets.
Table 4.
GSEA results of involution-specific signatures in IBC versus non-IBC in the merged 3 breast cancer data sets.
Table 5.
List of genes in Inv5 signature and genes enriched in IBC versus non-IBC.