How To Find Suitable Reference Genome After Genome Assembly

Question

[HG]

Hi everyone I have done 20 genome assembly of various strain of listeria monocytogenes**. The strain were collected from different species and different place. Now can any body suggest me how do i find a suitable reference for all 20 genome or else i have to take different reference every time for each genome??? and what will be best approach to find suitable reference genome ???

[DoubleDecker]

You mean you want to annotate your genomes now?

[HG]

Actually i did assembly using Spades, now i have scaffolding file so i want to see how this assembly and how much query coverage with reference genome

Answer 1

[Joseph Hughes]

You could automate a BLAST your 20 assemblies to known full genomes of Listeria monocytogenes. Use the blast results to work out the coverage of the closest genome in the database.

[HG]

I also thought like that but lets assume after assembly each of the genome have different number of contig and when you will blast it it will give you a separate result for each contig ?? so what will be suitable : shall i consider the largest contig of each assembly ??

[Joseph Hughes]

I would select it based on the total number of bases across all your contigs that are identical to a particular genome in the database.

Answer 2

[Alastair Kerr]

This information: http://bacteria.ensembl.org/info/about/species.html?search=Listeria+monocytogenes

Lists EGD-e in the taxonomic compara. This would be my 1st choice. However the other assemblies may also be useful for quality control mapping

[HG]

Could you please elaborate a little bit why you choose it ?? is there any rule ??

[Alastair Kerr]

Not really, just a generic answer to a generic question. The added utility of this genome is that it has been integrated into the comparative genomics analysis by ensembl

[HG]

If i want to make a genome analysis pipeline and i have to include it inside a pipe , so in that case how will i select ???

Answer 3

[Rohit]

If you are trying to annotate it then for the bacterial genomes its best to use the RAST.

http://rast.nmpdr.org/

[HG]

RAST has a problem identify of pseudogene. Moreover its not a reliable source. we have our in house annotation program i think its works better than RAST http://www.ncbi.nlm.nih.gov/pubmed/12682369

Answer 4

[SRKR]

I agree with Joseph, when you blast, you will get to know which genome shows the maximum similarity with all other genomes. That you should take as a reference genome rather than basing it on the size of the contig. Having a huge contig does not necessarily mean it will show maximum similarity with all other genomes.

[HG]

I agree with your opinion if it is single genome. But in my case i am doing 20 genome annotation simultaneously which are from different ecology and its a automation program that will take reads directly from machine after that it will do assemble and annotate. So you cant use single reference for all the genome. More over u think if you have around 10 contig for each genome you will get 10 blast result . Assume if two contig are very close in length but base is different : then it will give high identity and query coverage with different genome. Now it will be difficult to chose in a automation program. Hope you understand the problem now.