Figure 1.
Examples of visualization of gene expression patterns showing isoform variations.
A: Visualization of the expression pattern of MED24 gene showing an alternative start in 4T1. In the top panel, the horizontal scale corresponds to each probe set within the gene from the 5′ to 3′ ends. The blue bars indicate the comparison between 168FARN and 4T1 samples. From top to bottom we plotted the log2(fold-change) in expression, between the samples compared, and the statistical significance, −log10(p-value). The bottom panel shows the log10(expression intensity) of individual probe sets (from the top panel) in samples 168FARN and 4T1. Note that the seven last probe sets and the 3′ UTR region that are over-expressed only in 4T1 indicate an additional isoform in 4T1 starting from exon 20 start. B: Visualization of the expression pattern of SRRT gene showing an intron inclusion. In the top panel, the horizontal scale corresponds to each probe set within the gene from the 5′ to 3′ ends. The blue bars indicate the comparison between 168FARN and 4T1 samples. From top to bottom we plotted the the log2(fold-change) in gene-level normalized intensity between the samples compared and the statistical significance, −log10(p-value). The bottom panel shows the log10(gene-level normalized intensity) of individual probe sets (from the top panel) in samples 168FARN and 4T1. We note an intron inclusion between exons 5 and 6 in samples 168FARN. C: Visualization of the expression pattern of CD44 gene showing several internal cassette exons and three novel intron inclusions. In the top panel, the horizontal scale corresponds to each probe set within the gene from the 5′ to 3′ ends. The blue bars indicate the comparison between 4T07 and 4T1. From top to bottom we plotted the log2(fold-change) in expression, between the samples compared. The bottom panel shows a close-up of the CD44 region containing the differentially expressed exons and introns. The first custom track displays the fold-changes and the second custom track displays the sequencing alignment of the three retained introns. We note that in this example, exons 8, 11 and 13, two intronic sequences between exons 5 and 6, and one intronic sequence between exons 9 and 10, are over-expressed in the 4T1 sample.
Table 1.
List of some alternatively expressed probe sets.
Figure 2.
Proportions of tumor-specific over-expressed intronic regions.
The tumor samples are plotted on the x-axis, and the number of tumor-specific intron inclusion on the y-axis.
Table 2.
Top height over-represented biological functions and diseases for genes with isoform variations or whole-transcript expression differences.
Table 3.
Examples of significant genes, in the gene pathway, having important implications in normal biological processes and cancer.
Figure 3.
A network of molecular interactions containing differentially spliced or expressed genes between breast cancer tumors of varying metastatic phenotype.
Over- or under- expressed genes in 4T1 compared to 168FARN and 4T07 are respectively indicated by a green or a red color of the gene-product icon. The over- or under-expression rate is proportional to the color intensity. Genes that are not colored are those that are not differentially expressed or spliced in our data. The top functions or diseases where the gene-product are involved are cancer, tissue development, cell-to-cell signaling and interaction.