Only interchromosomal SVs used for chr_breakpoint_enrichment and pval_exp_chr ?

[e-sollier]

Hello,

I want to use ShatterSeek, and I tried testing it on a sample which seems (based on visual inspection) to have a chromothripsis event on chromosome 3. In the chromSummary that I get, most fields seem coherent to me, but for some reason I get NA for the fields chr_breakpoint_enrichment and pval_exp_chr for chr3, even though there are 11 intrachromosomal SVs on this chromosome.

I had a quick look at the code to try to understand what cause the issue, and it looks like only interchromosomal SVs are considered for these tests (https://github.com/parklab/ShatterSeek/blob/master/R/function_criteria.R#L264), even though, based on my understanding of the ’chromosomal breakpoint enrichment test’ and ’random distribution of breakpoints test’, intrachromosomal SVs should be taken into account for these tests.

Is this the intended behavior, or am I misunderstanding something ?

[isidroc]

Hello, thanks for your comment. When there is "chr" in the field it means that the calculation is applied to the entire chr, and not to the region encompassed in the SV cluster. The calculations at the chr level are informative sometimes to understand what is going on in the chr beyond the SV cluster. Does this clarify your question? Thanks

[e-sollier]

Hello, Thanks for your reply, but unfortunately it does not really answer my issue, so I'll try to reformulate.

I understand that for example for chr_breakpoint_enrichment, all of the breakpoints in the chromosome should be taken into account. However, I am under the impression that this is not what the code actually does. It seems to me that it only takes into account the INTERCHROMOSOMAL breakpoints for which one of the ends is on that chromosome, and ignores all of the intrachromosomal SVs (inversions, deletions, tandem duplications).

I have an example with 11 intrachromosomal SVs on one chromosome (3 DEL, 2 DUP, 3 2h2INV, 3 t2tINV). So it should be possible to compute chr_breakpoint_enrichment, but I actually get NA in the results for this chromosome. I looked at the code to understand why that is, and it seems that the value of chr_breakpoint_enrichment is only computed when there is at least one SV on this chromosome (https://github.com/parklab/ShatterSeek/blob/master/R/function_criteria.R#L269), but that only interchromosomal SVs are taken into account (https://github.com/parklab/ShatterSeek/blob/master/R/function_criteria.R#L264). This would explain why I do not get a value for chr_breakpoint_enrichment in my example, since I only have intrachromosomal SVs and no interchromosomal SVs, but this implementation does not seem to match the description of the chromosomal breakpoint enrichment test, since I can't see any reason why intrachromosomal breakpoints should be ignored here.

Do you see what I mean ?

[e-sollier]

Any updates ?

[alhafidzhamdan]

hii @isidroc, i have a similar problem with my dataset. NAs for chr_breakpoint_enrichment and pval_exp_chr even though visually it looks like chromothripsis. I wonder if there is any update to this ie https://github.com/parklab/ShatterSeek/blob/master/R/function_criteria.R#L269