If you know the name of the gene or it's DNA sequence, just run a BLAST search. If you have no idea what piece of DNA is associated with your trait of interest, then you have a LOT of work ahead of you. Start with a literature search.

In the past,

Genome of the bacteria of interest was mapped, then mutants lacking or having a phenotype of interest were complemented with fragments of DNA from a genomic library, then the gene was confirmed to be in that particular fragment from the genomic library, which restored the earlier phenotype of the mutant. further studies would locate the precise location of the gene in the genome.

Studies similar to above have provided us a wealth of information, the amount of which was at a point of time was very troublesome to manage and gave rise to the field of bioinformatics

With the information available to us in the literature and databases, one can obtain a homologous gene (with similar function or sequence) from another organism and obtain a fragment which can be used as a probe in southern blotting analysis of PFGE gels which gives a physical location of the gene on the genome.

NGS has revolutionized today's scenario,

the entire genomes of bacteria can be sequenced rapidly at reasonable prices, this combined with bioinformatics, you can map your gene a location on the genome

Nevertheless, few phenotypes can't be attributed to a single gene, or the gene responsible for a phenotype may not be identified yet. in such cases, classical techniques need to be used to identify the gene(s).

I agree with Vamsi.

First of all, you must know whether this organism have been sequenced before. In a positive case, you may find this information on the genome map or, at least, an indication of it.

If the genome have not been sequenced, an NGS approach to sequence the genome is the best way to achieve your purpose or to reduce the spectrum of research for future analyzes such as deletion assays.

If you know the name of the gene or it's DNA sequence, just run a BLAST search. If you have no idea what piece of DNA is associated with your trait of interest, then you have a LOT of work ahead of you. Start with a literature search.

I think you have to specify the bacterium you are interested in and you to the literature to see whether the gene is on a plasmid or not and you can choose the right or suitable approach to locate the gene,isolate or sequence it, depending on what is your aim. Literature is first.

There is an experimental solution to this, but I haven't got round to publishing it yet, so I'm reluctant to tell the world about it.

If you would like to contact me directly, we could perhaps arrange some sort of collaboration.

Sorry to be so mysterious, but it's one of those 'why didn't I think of that?' ideas.

I already did yesterday.

I was reading another discussion on researchgate and it turns out that I'm not the only one to think of this solution, so no need to keep it secret any more: here it is.

The principle is that plasmid-borne genes will be enriched relative to chromosomal genes in a plasmid prep.

Nonetheless, there will always be some chromosomal DNA in a plasmid prep and there will always be some plasmid DNA in a chromosomal or genomic prep.

What you do is make two DNA preps: a plasmid prep and a genomic prep.

You then run quantitative PCR for a chromosomal marker and the gene whose plasmid/chromosomal location you want to determine.

If both genes are on the chromosome, then their quantities will depend only on the amount of chromosomal DNA in each prep; their relative quantities will not change, although the absolute values will.

If one gene is on a plasmid, it's quantity in the plasmid prep will increase relative to the chromosomal gene.

In practice, you can just measure the Cq difference.

It's as simple as that. Note, however, that if the plasmid is very large (>200 Kb) you may need to use a method optimised for megaplasmids. Big plasmids are easily sheared, and if they lose their covalently closed circular form, they are indistinguishable from the chromosome.

An alternative (or complementary) method would be to do Southern blotting on a genomic prep without cutting with restriction enzymes. Your labelled gene probe would then hybridise to a band (or more likely several bands) running ahead of the chromosomal smear.

Andrew