I've done some transcriptomics work in the past with a polyploid organism, and this presented some unique challenges in the data processing and analysis. Since then, I have been brainstorming about the technical challenges one may face when sequencing and assembling the genomes of a polyploid organism. As far as I am aware, there are no polyploids whose genomes have been sequenced.
If one wanted to sequence, for example, a tetraploid organism, one approach would be to prepare and sequence all of the DNA together and then rely on post-sequencing analysis to tease apart the two co-resident genomes. However, it would be difficult, if not impossible, with this approach to distinguish inter-genome variation from intra-genome variation.
An alternative approach would be to isolate DNA from both co-resident genomes separately, and then sequence and assemble the genomes separately, so that inter-genome variation and homology need not be considered. However, I'm thinking at a very high level and have little intuition as to the technical feasibility of this approach. When there are two or more co-resident genomes, is it be possible to isolate DNA from only one of those genomes? What would this rely on (for example, would thorough cytogenetic/cytogenomic characterization help)? If this task is not possible, what types of limitations must be overcome to enable it?