I have a high-quality set of structural variant breakpoints from tumour/normal WGS data, and I am interested in digging into the various mechanisms that might be involved in each event.
There are several mechanisms that are proposed to play a role in generating structural variants, including Non-allelic homologous recombination (NAHR), non-homologous end joining (NHEJ/MMEJ), alternative end joining (Alt-EJ) and several replicative mechanisms such as Fork Stalling and Template Switching (FoSTeS/MMBIR).
Each of these are distinct processes, that have occur at different time points in the cell cycle and have different potential for generating genomic rearrangements. However, I have not been able to find a consistent requirement for each mechanism. I would like to know if anyone can think of a authoritative source for criteria I could use to form a working hypothesis as to the mechanism underlying each variant.
For example, if I have a 10 Kb deletion with 4 bases of microhomolgy at breakpoints I'm currently classifying it as an Alt-EJ event (largely based on the criteria I list below).
Several specific questions:
Microhomology at breakpoints is an essential requirement for several of these mechanisms (FoSTes/MMBIR). However, there appears to be no consensus as to whether this must occur exactly at the breakpoint, or whether it can be several bases away (for example here).
Secondly, according to this paper, the criterial used to classify structural variants is as follows (taken from the MeerKat manual):
NAHR: non-allellic homologous recombination, >100bp homology.
alt-EJ: alternative end joining, 3-100bp homology.
NHEJ: non-homologous end joining, 0-2bp homology or 1-10bp insertion at deletion break point.
FoSTeS: fork stalling and template switching, template switch, >10bp insertion at deletion break points.
However, almost every paper I read suggests different criteria, for example FoSTes events are normal associated with microhomologies at breakpoints rather than insertions.
Can anyone suggest a "consensus" of requirements for each mechanism?