Hi,

These are heatmaps I generated from "exotic" cutandrun experiments done at my lab.

The code I used looks like this :

plotHeatmap -m matrix.gz  \
     -out heatmap.png \
              --samplesLabel samples1 sample2 sample3 ... \
              --regionsLabel expressed_genes_fw expressed_genes_rv \
     --boxAroundHeatmaps no \
     --alpha 0.7 \
     --heatmapWidth 6 \
     --sortRegions keep \
     --colorMap Blues_r

What I don't understand is why the two first samples show absolutely no signal on the heatmaps but show high signal on the profiles. I understand when you choose to not set missing data to 0. My only guess is there are some pieces of signal not showing up on the heatmap that ended being used to compute the average profile.

I feel not setting missing data as 0 is better to have an overview of fragment length however I find the profiles a little missleading.

Any suggestions and comments are welcomed !