4. Interrogating Individual Samples

1. To interrogate individual samples you can plot their raw copy number values. In this example one may want to investigate what’s happening with the last normal sample (y2440_42) that exhibits copy number gains and losses across the entire genome.

   From the Normal vs Tumor Samples heat map view click the User Graphs node in the Graph Control Interface. Under Add Graph click the drop down and click Select Spreadsheet.... Choose the original MIP_Copy_Number Dataset - Mapped Sheet 1 spreadsheet where sample identifiers are column headers and the spreadsheet contains all copy number information. Click OK.

   Check the box for sample yw440_42_Copy Number and click Add.

2. The initial whole genome plot (Figure 16) appears to be an artifact of a very noisy sample with copy number gains and losses throughout the entire sample. (In Figure 16, the heat map has been hidden by unchecking the box in the Graph Control Interface to see a larger view of sample yw440_42’s data. The zoom has also been reset to a whole genome view by right-clicking the graph and selecting Reset Zoom.)

   

   Figure 16. Whole genome copy number data for sample yw440_42 displays noisy data

3. Applying median smoothing to the copy number values further displays a genomic wave effect. To do this, select the yw440_42_CopyNumber graph item in the Graph Control Interface and then click the Smoothing tab. Choose Median Smooth, Symmetric and a Window Radius of 5. Then zoom into any chromosome by selecting the chromosome number from the drop-down box in the tool bar. Chromosome 3 is a good example of the wave effect (Figure 17).

   

   Figure 17. Genomic wave effect as observed by median smoothing the reported copy number values