Pros and Cons of NextGen Sequencers

[ebhobbs]

Hi, I'm looking to find out the pros and cons of the each of the NextGen sequencers (Illumina Genome Analyzer, SOLiD, 454, Helicos). I'm especially interested in what sorts of experiments they would be best suited for. Also, if anyone knows of any papers that have been published on this subject, that would be great too.

[ECO]

Moving to general...

[krobison]

Here are some starting propositions which I'm happy to take flak on

Roche 454 - Longest but fewest reads per slide, fast runtime, most expensive per run. Still has trouble resolving long homopolymeric repeats. Long reads most useful for de novo assembly, SNP phasing, RNA isoform identification, so those are the sorts of apps which this platform will hold onto the longest.

Illumina - dominant platform, both in terms of installed base & bioinformatics developers efforts. Longest reads substantially shorter than 454 but longer than SOLiD; academic groups seem to be pretty much routinely doing 100+. Offers clever paired-end strategy, which is useful complement to mate-pair strategies (which tend to be profligate with input DNA & therefore unsuitable for projects lacking a surfeit of it)

SOLiD -- cheapest in cost per base right now. Read lengths may be topping out right now. Higher per-base accuracy due to color space encoding.

Helicos -- simplest sample prep, smallest sample requirements. Reads are short & have high frequency of short deletions. No amplification step should mean less bias and elimination of certain classes of errors (e.g. recent Nature paper suggested that reverse transcriptase errors are common in cDNA preparations).

Polonator -- the garage band sequencer. Cheapest instrument, lowest cost reagents, very short read lengths, fully open architecture for configuring your own assays & chemistries -- but very few users (but they are passionate).

ABI 3760. No, it's not next-gen but still holding onto some important niches. Still considered gold standard, still key on small projects or when you need to sequence specific clones.

[ebhobbs]

Thanks krobison for the quick reply. That's exactly what I was looking for.

[westerman]

Keith means the "ABI 3730" not the 3760 ... of course. We still get a fair amount of work done with our 3730 for the reason he mentions -- small projects, specific clones.

As for the SOLiD according to ABI reps its read length will stay at 50 bps but the coverage will increase several fold. Good for SNP and small InDel projects, at the very least.