Design and develop individual sample gene expression heatmaps

[logstar]

Design and develop OpenPedCan-api endpoints to transfer individual sample gene expression heatmaps and tables to MTP.

The formats of individual sample gene expression heatmaps:

• For each MTP disease page, the heatmap shows the expression of the most highly expressed all/PMTL genes in all cancer samples of the disease.
  • Rows: Most highly expressed all/PMTL genes. Limit to 30 most highly expressed all/PMTL genes by default. Rank gene expression by the average TPM values across all plotted cancer samples.
  • Columns: All cancer samples of the disease.
  • Colors: TPM or log10(TPM + 1).
  • Clusters: Hierarchical clustering of rows; hierarchical clustering of rows. Hierarchical clustering could be replaced by any other appropriate clustering method.
• For each MTP gene-disease evidence page, the heatmap shows the expression of the gene in all cancer samples of the disease.
  • Row(s): A single gene. This could be replaced with all isofroms of the gene, when the isoform data are available.
  • Columns: All cancer samples of the disease.
  • Colors: TPM or log10(TPM + 1).
  • Clusters: Hierarchical clustering of columns. Hierarchical clustering could be replaced by any other appropriate clustering method.

The development of individual sample gene expression heatmap endpoints could reuse @afarrel's plotting code in the differential gene expression heatmap endpoints, which is suggested by @taylordm.

Each table should contain the TPM values of all genes and samples plotted in each corresponding heatmap.

cc @chinwallaa