Feature request: grouped scaling

[SDeans0]

First, thanks for making a great tool. I have found it really useful.

I had an idea similar to (but distinct from) #219 which I think people would find particularly useful.

For RNA seq, it is common that you want to display expression at one locus in one condition but not another. If you define the scaling with auto, however, the conditions will be scaled independently, and you may see a large bar on the highest point of expression in the unexpressed condition, even if the absolute value is orders of magniture smaller than the expressed condition. Of course, it is possible to change this on an individual basis by setting a fixed scale, but the fixed value may not be appropriate for all of the regions in my batch.

Would it be possible to create a shared auto mode, where several tracks get scaled to the maximum and minimum across all of them? This is related to #219 because of the shared parameters, but different because the program still calculates them itself by observing the maximum and minimum values for each region.

[lldelisle]

This would be indeed a good add. However, I need to think how it can be done, for the moment, this is not obvious... I think about it...

[mtinti]

I had the same issue a few days ago. If it can help, this is my solution.

• compute the max/min values in the region of interest from the grouped tracks, outside of pyGenomeTracks (fast and easy especially for bigwig files).
• programmatically create a .ini file from a template where the first track of the grouped tracks set the max_value and the min_value with the precomputed max/min values in the region of interest.
• set all the grouped tracks in the .ini file (except the first) with "overlay_previous = share-y"

[lldelisle]

Thanks for sharing your solution.