With these results and given the limited data that is available to us, we cannot confidently establish the reasons why P. tricornutum lacks the methylation patterns that are observed in other very closely related diatoms. On the other hand, there are some interesting observations that we can make which can potentially lead to further investigations.

We can’t help but notice that P. tricornutum has a significantly lower median GC content (46.9%) compared to E. huxleyi for example. (65.7%). Furthermore, the difference in the fraction of methylated genomic cytosines is even higher, in fact higher than what the GC content difference alone can explain. The paper of Huff et al. \cite{Huff_2014} also explored the nucleus density of some diatoms, based on the fact that methylation patterns seem to drive nucleosome positions and hence the spatial arrangement of chromosomes. Unfortunately, they didn’t provide data for P. tricornutum and T. pseudonana. We have looked for the information in the literature and found estimates for both organisms \cite{Connolly_2008}. Even though the P. tricornutum strain for which we have the nucleus volume is different (CCMP1327 instead of CCMP632), we assume that it provides us with a relevant order of magnitude for the very least. We have added these data points to the plot from Huff. et al. (Figure \ref{fig:nuclei_density} in red and orange) and it is very clear that both organisms have much lower nuclear density than other related diatoms.