X-Shaped Pattern in Juicebox Visualization

[SEBASTIAN DONOSO]

Hello,

Could you help me understand what this pattern means? This is the Hi-C visualization of a plant genome. From what I understand, such strong and regular patterns are typically observed in smaller organisms, where the DNA is more compact.

Does this indicate that the assembly was done incorrectly?
Why is the X-shaped pattern so pronounced in this case?

On the other hand, if the assembly is correct, should I aim for this X pattern to appear more uniform during manual curation in Juicebox?

Thank you in advance for your help.

Best regards,
Sebastián

[SEBASTIAN DONOSO]

By the x-pattern, I’m referring to the interaction points that are distant from the main diagonal. Specifically, I mean patterns like those showing interactions between chromosome 1 and chromosomes 8 or 9, which shouldn’t be so prominent.

[Beisi Xu]

I am not an expert of Plant Genome though I find similiar pattern observed before for plant https://doi.org/10.1016/j.molp.2017.11.005 Fig1E Maize is not exactly like yours but maybe they could explained similiarly? Is your plant also transposable elements rich? https://doi.org/10.1038/s41594-024-01427-y

For my curiosity, are your MAPQ 30 version? If not, would q 30 show similiar pattern? The Mol Plant paper used MAPQ > 10, make me suspect mapping artifact. Does this plant's reference have well annotated centmere? 

[SEBASTIAN DONOSO]

Yes, the genome shown in the figure contains about 34% transposable elements, and from what I understand, this can help explain the X-shaped patterns. Transposons tend to accumulate in specific regions of the chromosome, and their spatial clustering, together with the Rabl configuration, contributes to the formation of such strong interaction patterns. Additionally, since transposons are especially enriched in centromeric regions, that could explain the strong signal observed in the central part of the plot, corresponding to interchromosomal interactions.

The heatmap I sent is unfiltered for MAPQ. However, the MAPQ >30 version shows the same patterns.

The current assembly does not yet include annotated centromeres.  

Thank you so much for sharing those papers they were exactly what I was looking for and helped clarify several aspects I had been trying to make sense of.