The Probe Trend Plot

The probe trend plot allows you to look at differences in read density over an "average" probe. It takes all the probes in a probe list and works out how many reads overlap with each position in each probe and then plots this out as an average over all of the probes.

For example, if you designed your probes over all the genes in your genome* then you could see if there was a general trend to having more reads nearer the 5' or 3' end overall.

The trend plot will plot values for probes with forward or unknown strand from left to right, and reverse probes from right to left so you can, for example, create probes over every promoter and get a sensible plot since the promoters from genes on the reverse strand will be adjusted to lie over the counts from those on the forward strand.

If your probes are all the same length (eg coming from a windowed probe generator or some feature generators) then the plot will be performed at 1bp resolution. If your probes are different lengths (eg from a contig generator) then each probe will be divided into 100 bins and reads will be mapped in that way.

A plot is drawn for every visible data store when you launch the probe trend plot.

Options

1. You can choose the type of plot. A relative distribution plot will weight each probe equally in the final profile, whereas the cumulative count plot will weight the probes according to the number of bases of read falling into each probe. The cumulative count plot is more susceptible to high read count outliers skewing your result, but will give you results in real read depths
2. You can choose which direction of reads to use for your counts
3. For the cumulative read count option you can correct your counts based on the total number of reads in each data store
4. You can scale each datastore so that you are looking at the variation in just that store without reference to any of the other stores plotted
5. You can choose to count each read position only once - removing duplicated reads from the plot
6. You can force a plot to be relative (uses 100 slots) rather than using bp resolution even if your probes are all the same length. This may make sense if you have very long probes
7. Once in the plot you can use the slider on the right to smooth your data using a running average over the raw counts

Once you have your plot you can choose to export either a view of the plot itself, or alternatively you can export the underlying data as a tab delimited text file.