I am currently tasked with assembling a transcriptome for a plant species.

There is a published genome available for this species, which is around 70% complete.

Is it better to do a genome-guided assembly (say, using Trinity) using this partial genome, or to do a de-novo?

Is there another approach that may improve the genome - e.g. merging this 70% genome with the published genome of a similar species?

Are there any other specific issues that should be addressed (and if possible, how) when doing a genome-guided assembly with only a partial genome?

Thanks in advance