Atypical holocentromeres as in C. elegans

[V-JJ]

Hello!

Thanks for the software, it looks great!

I have some questions regarding the use of RepeatObserver in Holocentromeric species such as C. elegans. Have you tried RepeatObserver on it? Since it has diffuse holocentromeres, it's quite tricky to get to their description. As fas as I know this species doesn't have a clearly defined centromeres (delimited by tandem repeat or TE).

Thanks in advance,
V

[jonnyPizza]

Hi,
I have the same question about whether or not this can be applied to holocentirc chromosomes as well such as Rhynchospora pubera (beak-sedge). I do know however in this species, the centromeric regions are closely associated with a Tyba repeat. In my experience, centromeric regions are composed of repeated sequences but are much less dense/compact and lengthy as compared to monocentric species.
I will let you know how my tests go. Please let me know if you've had any success.

Best,
J

[V-JJ]

Hi @jonnyPizza !

No success with C. elegans. But, regarding Rhynchospora:

• Fourier heatmaps don't show anything clear.
• However, if you get the raw Shannon diversity plots first and you zoom in (capture below for example), you may observe some regular drops in Shannon diversity index that could be (probably) the regions with Tyba. If you do that, note that in IGV (image below) you should play with scale (upper left corner) to discard some extreme values.

Overall, I think that the results of RepeatObserver are very promising and interesting to explore. However, it is not straightforward.

• It seems to work only on species with repeat-based centromeres and the results depend on the genome itself. For instance, in Luzula sylvatica it worked much better (it was present in their biorxiv).

• Additionally, you need to dig through the data (zoom in, plot the results by yourself, discard some outliers) to be able to see some patterns (if there is any of course).

• I plotted in IGV the information from the following files, located in Summary_output/output_data
  -- *H0-AT_Chr1_Repeat_abundance_sum.txt
  -- *_H0-AT_Chr1_Shannon_div.txt

Also, if you suspect/know that your species has repeat-based centromeres, you could add TRASH results (for tandem repeats) or RepeatMasker data, etc... and explore evidence from several sources at once.

Thanks,
V

[celphin]

Hi V-JJ and jonnyPizza,

Sorry for the very slow response. I was swamped defending my thesis and then was away traveling for a bit.

I have not tried the program on that many holocentric chromosomes so thanks for looking into this more!

I ran Morus notabilis (mulberry) (Figure 3E), Luzula sylvatica, and Chionographis japonica (Figure 8 C and D). In all of these plant genomes with holocentric chromosomes, the multiple centromere locations were obvious. I had not tried it in C. elegans though so it is interesting that there are no clear repeats.

It is definitely true that in some species with large chromosomes seeing the repeats in the spectra or Shannon diversity plots can be hard and you may have to look in more detail at the output like V-JJ suggests (e.g. IGV). I would love to make a Zoom-able plot in R to look at the chromosome spectra in more detail but have not found a good way to do this yet. You can determine all genome positions in the Shannon diversity or repeat abundance files that are more than one or two stdev above or below the mean and these indicate the likely centromere locations - although you may need to adjust the number of stdev for holocentric chromosomes with many repetitive regions throughout.