The HiC Heatmap Plot

The HiC heatmap plot is a global way of being able to look at the strengths of interaction between different regions of the genome. It is only available when the currently active DataStore is a HiC dataset.

You access this plot by selecting "Plots > HiC heatmap" you can then choose whether the axes for the plot are shown as genomic coordinates, or as sets of probe lists which you can then select.

The plot shows two copies of the genome at right angles to each other and between these it draws a heatmap showing the density of paired reads which fall into each interacting pair of regions in the plot. The potentially interacting regions are taken from the currently active probe list.

To zoom into the plot you can simply drag a box within the main plot area. To zoom out press the right mouse button. If you hold the control key while zooming out only the x axis will zoom out. If you hold down shift then only the y-axis will zoom out.

If you double click on any interaction point then one end of the interaction will be shown in the main chromosome view. If you shift+double click then the other end will be shown.

Options

There are a number of different filtering options which apply to this plot. You can apply most of the filters either before the plot is calculated or after it has been displayed. The filters you apply before calculating the plot are hard limits which allow the program to immediately discard many of the potential interactions to leave a manageable subset of interactions to display. Many of the same options appear after the plot has been drawn and these can be changed interactively, but at this stage the filters can only be set to values more restrictive than those initially chosen when the plot what calculated.

Pre-calculation options

• You can choose which dataset and which probe list to use for the plot. The dataset must have originally been imported as a HiC dataset.
• You can set a minimum and maximum range of distances over which interactions will be considered. Entering no value removes the filter. If any value is set for the maximum distance then all trans interactions will be discarded.
• You can select the level of enrichment you want to see in your interactions. Currently enrichment levels below 1 cannot be selected
• You can select a significance cutoff above which interactions will not be shown
• You can set a minumum absolute number of observations which are required for an interaction to be shown - even if it passes the enrichment and significance filters
• You can choose whether the statisical model corrects the expected values for physical linkage distance when considering cis interactions

Post-calculation options

Once the plot has been drawn you can adjust all of the values set in the pre-calculation filters. In addition you have some more options.

• You can use the 'Match Chr View' button to set the HiC plot to the exact region current visible in the chromosme view.
• When the region shown in the plot on the X or Y axis falls entirely within the same chromosome you can use the "Send X" or "Send Y" buttons to update the main SeqMonk chromosome view to the region currently shown on the HiC plot X or Y axis.
• You can use the "Make Report" button to create an annotated report of the interactions which pass the current set of filters
• You can change the parameters used to colour your plot. By default this will be the obs/exp value, but you can use the p-value, number of interactions or the value associated with the underlying probe in the current quantitation.
• You can choose to filter the interactions you're seeing to show only interactions where at least one end matches a probe list which you select
• You can use the "Cluster interactions" button to perform a hierarchical clustering of all of the interactions to bring together sets of probes with similar interaction patterns across the whole genome. Once this clustering has been performed you will then have the option to change the clustering threshold to generate either tight or loosely coupled clusters. You can also choose to save your clusters back to your original project as probe lists.