Composite profile plots. To contextualise how the top scoring hsa-miR-124 hexameric seed query (i.e. GCCTTA) enrichment profile compared to that of all other unique nucleotide hexamers, an equivalent analysis was completed using SBSE with each of 4096 (i.e. 4^6) unique hexameric seed queries. The resulting information is then intuitively represented by composite plots of the, previously described, enrichment score. The composite plots succinctly summarise the estimated enrichment scoring of all 4096 unique hexameric (seed) queries. Each of the four PanelsA-D represent the analysis of a separate time-point following hsa-miR-124 transfection. In each instance the x-axis represents the ranked transcripts (i.e. by fold change, from most up-regulated (left) to most down-regulated (right)). The x-axes scale indicates the number of bins used in the analysis (See methods for further detail of bin implementation). The y-axis represents the enrichment score which is scaled dependent on the range of enrichment scores encountered within the dataset. Each of grey lines represents the estimated scoring of a unique hexameric query sequence, while the highest scoring hexamer at each time point is coloured turquoise. The hsa-miR-124 GCCTTA seed motif is coloured red throughout and the maximum enrichment score "i*" is indicated on PanelsB and C. Furthermore, in Panels B and C the maximum enrichment score is the query sequence. The selected data clearly summarises how the hsa-miR-124 motif gains in prominence in each of the post-transfection samples and becomes undetectable if the differential expression profile is shuffled (Panel D). In particular, note how the overall enrichment score of the datasets fluctuates post-transfection. Initially (Panel A) all data forms a homogeneous body with no enrichment score above 40. 16 hours post-transfection (Panel B) a large number of up-regulated AT-rich transcripts are obvious (indicated by the blue arrow). After 24 hour post-transfection (Panel C) this collection of up-regulated transcripts are no longer apparent.
Wilson and Plucinski BMC Genomics 2011 12:250 doi:10.1186/1471-2164-12-250