Comments on revised manuscript

I read through the revised article and they addressed some of the comments I made. My general comment, which they have not addressed satisfactorily (for me) is still the pseudochromosome assembly method. From the start, only one method for reads de novo assembly was used. From our rice genome experience and current experience in tree genomes, we commonly use >1 to nominate a best method, or even combine outputs of the methods for improved scaffolding results. As you can see the N50 of their assembly is fairly small (187kb), it’s not convincing for a claim of high quality reference genome. Anchoring the scaffolds by alignment to a known high quality reference for asserting scaffold ordering & pseudochromosome assembly, while this may be acceptable for animal genomes, is not enough for plant genomes. There’s a need to show that their particular cannabis accession (cb2) was most closely related to cs10, not to PK. This could easily be done with their reads and contig-level assembly data. By asserting scaffold order /anchoring to cs10 (which is a very different accession to cb2, as the authors pointed out in terms of chemotype), all subsequent cb2 genome comparison results is being ascertained with cs10 information. Without cb2-specific information (genetic map, optical mapping), the pseudochromosome assembly is not so solid. Here’s a template paper they can refer to … https://www.nature.com/articles/s41597-020-0438-2 For me, it would be more proper to report the genome as a draft assembly of 8,477 contigs, then include a contigs tiling path table vs cs10 or pk (depending on which accession cb2 is most closely related to). All the rest of the analyses on genome annotation would remain as-is, and is an important resource. Additionally, reports on the repeat content (TEs, etc) to which they did the needed analysis already, genome heterozygosity /haplotigs would be very interesting findings to see, and the authors could easily do / may have done these analyses already, and just need to present these. Asserting low heterozygosity by casual observation is not enough, they have the data to measure this and report accordingly.