search

eblerjana / long-read-1kg

Code written for the long-read 1KG project
MIT License
0 stars 1 forks source link

concerns about graph augmentation #1

Open coopsor opened 2 weeks ago

coopsor commented 2 weeks ago

I've been closely following your recent paper on graph augmentation, which I find quite intriguing. However, I have a few questions. Could you elaborate on the advantages of choosing to augment a pangenome graph based on structural variations identified through comparison, versus using an assembly-based augmentation approach? Additionally, have you conducted any comparative experiments in this regard?

eblerjana commented 2 weeks ago

We augmented the graph with SV calls made across ~1000 genomes sequenced with low coverage ONT reads. The reason why we did not augment it with assemblies is simply because that would have required generating (haplotype-resolved) assemblies for all ~1000 genomes, which was not possible with sufficient quality due to the limitation in coverage and even if, would have been computationally very expensive to do. If assemblies were available, this would most likely be the better approach