MiSeq 600bp sequencing problem

[LizBent]

Hi, our Genomics Center sent me a link to the MiSeq run they are doing for us with the new 600 bp chemistry - and I notice something on the run stats that scares me:

1. Yield perfect read 1: 170.8M
2. Yield perfect read 4 (second 300 bp read): 9.6M!

The run is still going- will the second read keep going up? It started at 8.9 M for this statistic, and I panicked because in the first attempt to sequence my amplicons, the second read was a lot lower than the first.

I'm worried because I added indexes and adapters via PCR (regular primer, then adapters+ regular primers, then indexes+adapters in 3 sequential PCRs) and I'm so afraid I screwed that up.

[TonyBrooks]

Hi, our Genomics Center sent me a link to the MiSeq run they are doing for us with the new 600 bp chemistry - and I notice something on the run stats that scares me:

1. Yield perfect read 1: 170.8M
2. Yield perfect read 4 (second 300 bp read): 9.6M!

The run is still going- will the second read keep going up? It started at 8.9 M for this statistic, and I panicked because in the first attempt to sequence my amplicons, the second read was a lot lower than the first.

I'm worried because I added indexes and adapters via PCR (regular primer, then adapters+ regular primers, then indexes+adapters in 3 sequential PCRs) and I'm so afraid I screwed that up.

The indexes and adapter sequences should be on the same oligo, so if you messed up the indexing PCR, you'd fail to cluster. What's your cluster density?

[LizBent]

the indexes and adapter sequences should be on the same oligo, so if you messed up the indexing pcr, you'd fail to cluster. What's your cluster density?

1067 +/- 23

[GenoMax]

Is this a 2D barcode or 1D barcode run? V2 or V3 chemistry?

If this is V2 chemistry and these are amplicons then this run is overloaded at 1000+ clusters/mm^2.

Are you sure you have 170M reads for Read 1?

[TonyBrooks]

Optimal loading for amplicon on V2 is around 800K/mm2, but it's very library specific.
We've had libraries run at 1000k/mm2 with no problem at all.

Either way, it looks like you library is clustering. Next thing to check is the Phas/Prephas. You're looking for values below 0.1% for your reads, indexes are usually higher. Higher values indicate a possible problem with your reagents.

The % perfect applies only to the PhiX spiked into the run, hence the low yield.

[AllSeq]

Are you sure you have 170M reads for Read 1?

Same question: are you sure you have 170M reads? The spec is 25M.

[TonyBrooks]

Same question: are you sure you have 170M reads? The spec is 25M.

Pretty sure you're looking at the "Yield Perfect" column in BaseSpace, which is the PhiX reads. M = Mb here, not number of reads
My last run had 12.78M reads pass filter = 3.2Gb for R1 (250bp * 12.78M)
I has 10% PhiX spiked in, so 320Mb of that 3.2Gb was PhiX
Yield Perfect was 240M, so I had ~75% %Pefect (in actual fact, it was 75.4% according to BaseSpace)

[LizBent]

The run statistics are getting better, so I think I was freaking out for no reason... I have 3.4G of data from read 1, and it's halfway through read 2 and there are 1.7G of data, so this run should (?) be fine...

Sorry to have bothered you all. I really hate the BaseSpace run summary interface!

[GenoMax]

At times like these having less (or not continuously updated) information actually turns out to be better :-)

[microgirl123]

At times like these having less (or not continuously updated) information actually turns out to be better :-)

So true. I just started my first V3 run Tuesday so I was watching it closely on Basespace. First up pops the cluster density at 523 +/- 477! Panic inducing! Turns out to be 983, which is perfect. Then clusters passing filter popped up at 43% +/- 52%! Completely useless information but worrisome. Turns out to be 91%, so I'm happy.