titanium vs standard chemistry for BAC assembly

[Hilary]

Hi,
I am planning to have a couple of BAC clones sequenced on a 454 and it looks like I have a choice between standard chemistry with a paired-end library, or titanium chemistry with no paired-ends. Does anyone have any experience with which option is likely to provide the better assembly? Does the longer read length with titanium chemistry out-weigh the disadvantage of not using paired-ends? The BAC clones are likely to have some regions of repetitive sequences, and they are from a reptile with no close relatives so there is no reference.
-Hilary

[kmcarr]

Hilary,

My choice would be for the standard paired-end run. The additional read length of Titanium is unlikely to help span repetitive regions unless their lengths happens to fall in the range between the average read lengths of standard vs. Titanium (250 vs 450nt). I'm not saying that the paired-ends will close these gaps, they likely will not, but at least you will be able to orient the contigs. The other advantage of Titanium vs. standard, the amount of raw data, ~500Mbp vs ~100Mbp, is also not really useful in your case. You didn't say how big your BAC is but assuming it's a typical BAC even the standard chemistry is going to give you more coverage than you need. (In fact the assembly will likely be helped if you limit the amount of data you use by randomly selecting a subset of the raw reads; start by using a set of data equivalent to ~30X coverage.)

As an aside, Roche has just released the paired-end chemistry for Titanium (at least in the USA, don't know about NZ.) However they also strongly encourage users to attend training before working with the new kits so that may be a hold up as well.

[Hilary]

Hi,
Our 454 facility isn't offering titanium paired ends commercially yet, and I don't know that my budget will stretch to it anyway, so I think I'll go with the standard paired ends for now.
Thanks for the advice!
Hilary