Scan see that the E2 region has a lot of predictive power. as evidenced by the thick column of predictions in the third group (it may only look like the second but Core is very small). The P7 region also has quite a bit of predictive power. This is currently in "alphabetical order" of the genes ... when I re-order it:
Things make much more sense. Now the predictive regions are grouped together.
The circos diagram shows the same basic results except in a more visually appealing way. However, I'm having trouble posting it here ... probably a size issue. Despite converting it to a JPG at 5% quality its still over 2 mb.
The interesting thing will be what is the position which has that high prediction level. So in this poorly labeled histogram ... its just a quick one I'll do a better one tomorrow, just want a general idea:
You could guess that the E2 position: 176 (alignment-space) is that large spike and the second largest is at E1 475 (also alignment-space) ... I will have to do a little work tomorrow to convert these to sequence space.
I'm going to spend some time tomorrow getting more information into the above figure ... perhaps using the "color" of each scatter point indicating the level of linkage ... the other strategy would be to use a pcolor plot. Also doing joint histogram/pcolor would be very helpful.
No comments:
Post a Comment