Getting started with de novo genome sequencing

[michaelbarton]

I'm new to genome sequencing and I'd like to ask for some general advice. We'd like to sequence as many P. fluorescens (Pf) strains as we can manage whilst try to get a reasonable assembly. We're using 454 as these are the facilities we have available to us. The Pf genome is ~7Mbp and does contain regions of IS elements which could make assembly more difficult. Moreover existing Pf genomes show low sequence similarity ~60% which, we assume, makes them not very useful for comparative assembly.

Our main concern is how much genome coverage we need for an adequate assembly. I realise this may be a relative measure that varies from organism to organism but we are considering 15X coverage as this allows us to get four genomes from a 454 plate. However it has been suggested to us that from previous Pf sequencing this will not be enough to completely assemble given the repetitive elements.

Any thoughts or suggestions are welcome. We are a small lab and trying to work things out as we go along.

[flxlex]

I agree that 15 x coverage shotgun reads will not give a very high quality assembly (although lots of information can be gained from it).

According to Roche/454, a single plate divided in four lanes can actually give a very good assembly for four genomes. You could get one scaffold per genome (consisting of contigs spaced by gaps).
However, it requires making 8kb paired end libraries. If you are able to do that, or have it done for you, the additional cost (and time) should well be worth it.

Check out the pdf:

http://www.454.com/downloads/DeNovoSmallGenomes_Flyer.pdf

flxlex

[michaelbarton]

Thanks for the suggestion, we will try to get paired end reads as hopefully this should improve the assembly.