[penpen]

here's one paper of combining application of GAII x, FLX and ion
torrent. sequencing by GA IIx , FLX, and SNP validation(amplicon sequencing) by ion torrent......

[dongzw]

Very good question! And l'd like to see some reviews on this topic

[rnaseek]

Best Benchtop sequencer: MiSeq or PGM or 454 GS Junior?

Finally a performance comparison of MiSeq, PGM and 454 GS Junior by researchers not the vendors

http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2198.html

[snetmcom]

Finally a performance comparison of MiSeq, PGM and 454 GS Junior by researchers not the vendors

http://www.nature.com/nbt/journal/va.../nbt.2198.html

expect the Miseq and 454 runs were performed by the vendors.

[nickloman]

Just to be completely accurate, we also performed the 454 Jr runs. When we did this study the MiSeq hadn't yet been released.

[Nitrogen-DNE-sulfer]

Great paper Nick. Does anyone have experience with Low Amplification libraries on the MiSeq? Their protocol demands 10ul of library be diluted into 1ml of HT1 and this is further diluted 2:1 before loading so a 200 fold dilution to load. Best I can tell the MiSeq needs 300-600ul in machine for the cluster formation (mostly to prime the lines). I bet only 50ul are actually in the chip during clustering so most of this sample is lost. As a result we are having a hard time delivering less than 18 cycle haloplex to the instrument and we can know this many cycles is not good for even coverage.

[pmiguel]

Great paper Nick. Does anyone have experience with Low Amplification libraries on the MiSeq? Their protocol demands 10ul of library be diluted into 1ml of HT1 and this is further diluted 2:1 before loading so a 200 fold dilution to load. Best I can tell the MiSeq needs 300-600ul in machine for the cluster formation (mostly to prime the lines). I bet only 50ul are actually in the chip during clustering so most of this sample is lost. As a result we are having a hard time delivering less than 18 cycle haloplex to the instrument and we can know this many cycles is not good for even coverage.

I am with you on this. The cBot/HiSeq seems to waste quite a bit of library, but the MiSeq is even more extravagant.

That said, obviously you could skip the the final 2:1 dilution down to 10 pM, by starting at 1 nM, instead of 2 nM. And, since you only use 600 of the 1 ml of neutralized ssDNA, you could start with 6 ul of library, instead of 10 ul. Together that would allow you to back off 2 cycles on your PCR.

That said, there is still the issue that 18 cycles yields a theoretical maximum of 2^18X amplification. ~256,000X to get to 2 nM. That suggests your initial concentration of amplification-competent library molecules is on the order of 0.01 pM. That would be about 10-20 million library molecules/ul--ie, probably below 10 pg/ul. Is that what you expect? If not, it could be library construction or enrichment PCR needs to be optimized a bit.

--
Phillip

[ECO]

Great paper Nick. Does anyone have experience with Low Amplification libraries on the MiSeq? Their protocol demands 10ul of library be diluted into 1ml of HT1 and this is further diluted 2:1 before loading so a 200 fold dilution to load. Best I can tell the MiSeq needs 300-600ul in machine for the cluster formation (mostly to prime the lines). I bet only 50ul are actually in the chip during clustering so most of this sample is lost. As a result we are having a hard time delivering less than 18 cycle haloplex to the instrument and we can know this many cycles is not good for even coverage.

The instrument actually takes 400 out of the 600 in the well as best I can tell from reading the recipe files...unfortunately the post run cleanup washes flush out that well so it's hard to estimate the dead volume...but I bet you could get away with less than 600.

For denaturing low concentration libraries (and all libraries for that matter), I've moved 100% to using NaOH followed by neutralization with HCl, followed by dilution with HT1. This removes the (somewhat ridiculous) need to dilute out the NaOH...and frees you to denature any way you want.

My preferred protocol is to dilute the library into 40ul of water or EB, add 1ul of 2N NaOH (to get to 50mM), incubate 5 min, and add 1ul of 2N HCl. then dilute this directly to whatever loading concentration I want....this works great and prevents having to denature a huge quantity of library.

Next on my list is to add ~10mM Tris to the HT1 to further smooth out any variations in pH (I think this is recommended in the Sanger optimization paper), but I haven't run into any problems yet.

[Diagnostics Joe]

Pacbio is worth considering too. However if you are doing lots of MB/whole genome sequencing then Illumina is the most cost effective option these days? I've done a summary of the pros/cons here