How to comparing two bacterial genomes and find the mutation, rearrangement etc.

Question

[FAST_GENOME]

Dear All, we have two different strains of bacteria. One has biocide resistance and another one doesn't. We try to find out which and how many mutations are involved in the resistance. I am thinking of sequencing the 2 strains and compare each other. I am not sure if this is a correct idea. Can someone give some pointers? And, regarding the comparison and variant identification, what tools I can use for this purpose?

Appreciate it in advance.

Answer 1

[Joe]

Yep. Sequence them both.

If it's a bacteria without an existing reference genome, maybe consider long and short read technologies so that you can finish the genomes.

If you think the resistance might be down to SNPs rather than the presence/absence of particular genes, you'll need to ensure you get decent quality sequencing and at a reliable depth (probably at least 30-fold).

What tools to use for the comparison largely depends on the types of differences you're looking for.

If you're interested in SNPs, you'll want to align all the reads for one genome against the other, and then use a tool like VarScan or Breseq to call the variant nucleotides.

If you're interested in finding missing/gained genes, you can align the genomes with tools like MUMmer and visualise it in many different ways.

[FAST_GENOME]

Thank you so much. I just find an available genome on NCBI. Can I map the reads from two strains against the genome ref and filter the mutation only exist in the biocide resistance strain? Thanks.

[Joe]

Yep, that's one way you can do it.

A word of caution here though, the sequence deposited at NCBI might be quite different from yours, even if they're the same bug, due to passage in lab conditions. You may find more variants than you expect which make finding your relevant ones a little tricker. This is why I suggested sequencing your "Wildtype" and your resistant "mutant", which may do away with many of these problems.

It depends very much on the organism though. No harm in trying it anyway if you\ve got the reads from both (it's often interesting to know how freakish your lab strains are anyway).

[FAST_GENOME]

Appreciate your help.