I have paired-end Illumina data of a bacteria (1) and a reference genome of the same bacteria (2). However, the reference bacteria is not the same as (1), because (1) was treated differently and changed it`s genome. Now I want to assemble the genome of (1) using the reference genome, since de novo assembly with only Illumina reads produces a lot of contigs.

I mapped the reads of (1) to the reference genome and got 99% Pairwise identity and 35.9% Identical sites. What is the best way to get the correct genome of (1)?

I could modify the reference genome until it fits all the reads. Regions of low/high coverage indicate, that the reference genome is the wrong representation of (1) at this spot. To correct this I could try to find contigs by de novo assembly of (1) that explain this area better.

This is my idea, but I want to ask you experiences people, how you would approach this problem and what tools I may use.