In what case there are RR reads but no LL reads when detect inversions using IGV?

Question

[Jingfang]

Hi all, From Integrative Genomics Viewer User Guide, when an inversion shows up in paired-end reads, there should be both RR reads and LL reads. As shown below. When I use the orientation of Illumina paired reads to detect inversions, I found a region only have RR reads, see the picture below. Looking forward to getting help.

[Istvan Albert]

you should visualize the image as pairs (view-as-pairs) the image you show is not quite right, it is also important to note what gets colored,

also I would say the explanations by the IGV manual appear to be confusing. and inversion would do two things: make both ends of the pair align in the same direction and make the insert much longer.

as a matter of fact I don't think their example image is correct, all reads starting right at the border of the inversion point should have their mates further away while pointing the same way.

if it is a true inversion then it makes no sense that some read pairs over the junction still align correctly.

I will say that figuring out structural variation from IGV image alone is a tricky business.

[Jingfang]

Thanks for your reply.

The IGV image I show used color alignments by pair orientation, and view mate region in split screen. Blue reads in two screen were in pairs.

I agree with your two points about inversion. There will be correctly aligned reads near the junction, but no reads cross the junction.

[Istvan Albert]

I think a view as pairs is much cleaner as it shows the actual pairing and how they cross the gap. In the view you show it is not clear where do the pairs start and end.

Here I would say that if there is an inversion it is in addition to the existing region, and not in place of it. This is why so many reads do not show the variation.

On top of that you seem to have some sort of copy number variation which lines up with having an inversion in addition.

Still does not explain where the other inverse reads went, though I am not sure how your reads are colored. IGV can be very confusing actually.