[GenoMax]

Yes, that is the question: does SBS v4 + patterned flowcell cluster kit = running on a HiSeq "1T" == a HiSeq X Ten single instrument?

We are assuming that there will be a "patterned" flowcell kit for 2500. I found this blog post (skip to the "Edit" at the bottom of the post), which says that "patterned" flowcells will only work with HiSeq X and TruSeq Nano kits.

Let us hope that is not true.

[pmiguel]

Following is pure speculation so take it with a grain of salt.

I expect to see patterned flowcells trickle down to the 2500's (and 1500?) at some point in time this year (this is the upgrade most of us are waiting for) though it remains to be seen if a "new camera requirement" will go with it, which could leave a number of machines out in the cold. It is difficult to tell (from pictures) if the patterned flow cells are the same dimension as the current flowcells. If they are not then that could be yet another problem (unless illumina makes a new version of patterned flowcells for existing machines).

We know that 1T machines will use a new version of SBS reagents (v.4) along with some new software/firmware. Is that kit anything like what the X will use, anyone?

Yes, that is the question: does SBS v4 + patterned flowcell cluster kit = running on a HiSeq "1T" == a HiSeq X Ten single instrument?

v4 (if your instrument can run it) gets you to $28/billion bases (reagent costs). Does the patterned flowcell take you to $7/billion bases (reagent costs)?

In other words, say I am a late Cretaceous dinosaur watching a light in the sky heading for Chicxulub. Am I avian, or non-avian?

--
Phillip

[GW_OK]

http://www.illumina.com/systems/hise...fications.ilmn

v4 is mentioned here for the 1500/2500's. Looks like a higher cluster density and a corresponding bump in output (up to 2 billion reads per flowcell).

From the fine print on the SBS v4 page:

*Compatible with factory-built HiSeq 2500 and 1500 units and select HiSeq 2000 units. Please contact Customer Service to confirm system compatibility.

So v4 is Hiseq1T chemistry only.

[TonyBrooks]

Following is pure speculation so take it with a grain of salt.

I expect to see patterned flowcells trickle down to the 2500's (and 1500?) at some point in time this year (this is the upgrade most of us are waiting for) though it remains to be seen if a "new camera requirement" will go with it, which could leave a number of machines out in the cold. It is difficult to tell (from pictures) if the patterned flow cells are the same dimension as the current flowcells. If they are not then that could be yet another problem (unless illumina makes a new version of patterned flowcells for existing machines).

We know that 1T machines will use a new version of SBS reagents (v.4) along with some new software/hardware. Is that kit anything like what the X will use, anyone?

http://www.illumina.com/systems/hiseq_2500_1500/performance_specifications.ilmn

v4 is mentioned here for the 1500/2500's. Looks like a higher cluster density and a corresponding bump in output (up to 2 billion reads per flowcell).

[GenoMax]

Following is pure speculation so take it with a grain of salt.

I expect to see patterned flowcells trickle down to the 2500's (and 1500?) at some point in time this year (this is the upgrade most of us are waiting for) though it remains to be seen if a "new camera requirement" will go with it, which could leave a number of machines out in the cold. It is difficult to tell (from pictures) if the patterned flow cells are the same dimension as the current flowcells. If they are not then that could be yet another problem (unless illumina makes a new version of patterned flowcells for existing machines).

We know that 1T machines will use a new version of SBS reagents (v.4) along with some new software/firmware. Is that kit anything like what the X will use, anyone?

[pmiguel]

Anyone want to comment on the differences between a single HiSeq X Ten and a HiSeq 2500 1T?

There isn't much point trying to run a 1T lab -- same type of data for 4x the price. Unless it is just a matter of time before 1Ts will be running X Ten chemistry.

--
Phillip

[ericchow]

According to the NextSeq spec sheet (http://res.illumina.com/documents/pr...nextseq500.pdf), in addition to the 520 and 650nm LEDs, it also has a 780nm laser. At first I thought this might be for the reagent scanner, but the MiSeq spec sheet doesn't list it and the HiSeq spec sheet lists the scanner as 650nm.

Could this infrared laser be used to detect all clusters?

[pmiguel]

Well, as Genomax has posted from the user guide, the first 2 bases are the critical cluster calling cycles. I'm going to go find some data and count the number of reads with GG in the start. I will wager that it's a low percentage.

But, to paraphrase Gordon Gecko: Cynicism is good.

What? Does the NextGen do clustering on the index read? (The SOLiD actually did. EDIT: wait, it did the index read first, but it identified clusters (beads actually) using a scan from a fluor annealed to all the beads) If not, then the task is much simpler -- just find a previously identified cluster and call bases. The fact that the NextGen can't handle 2 initial G's in the index read, suggests it is going to be much less tolerant of G's during clustering.

--
Phillip

[GW_OK]

Dear Illumina,
Please release a high-GC content data set from the NextSeq to BaseSpace.

Thanks,
SeqAnswers readership

[jwfoley]

The reps in the webinar just said a non-G is required in the first five bases of the read, but for simple statistical reasons that's not likely to be much of a problem.

However, I worry about mild bias against reads that are G-rich in their whole length, not just severe bias against ones that are G-rich at the beginning. This could be particularly troublesome for specific applications like CpG methylation profiling.

[GW_OK]

Well, as Genomax has posted from the user guide, the first 2 bases are the critical cluster calling cycles. I'm going to go find some data and count the number of reads with GG in the start. I will wager that it's a low percentage.

But, to paraphrase Gordon Gecko: Cynicism is good.

[pmiguel]

Yeah, that'd mean an entirely new, machine-specific library kit and backwards compatibility issues.

Is this not just a little bit nit-picky, though? Think of a genome. Think of an amplicon. How many times would arise wherein the first X bases were a G-homopolymer? Theoretically it could happen, certainly, but practically, I'm thinking not so much.

Anyone have any concrete examples of G-heavy read starts?

Well the stranded TruSeq RNA prep libraries tend to be quite G-rich in their early bases.

Call me cynical, but I just know this is going to cause bias unless Illumina has substantially altered its cluster calling software. I am guessing anything with 3 or more G's in its read1 first 5 bases is going to be "lost" by the software as a cluster a large percentage of the time.

Not that Illumina will have any trouble competing with the Proton.

Also, after a year or so Illumina will probably fix the issue. That is my prediction. Feel free to ridicule me if it didn't go down that way 1 year from now.

--
Phillip

[jwfoley]

Here are some prices I've heard:
FC-404-1002 | NextSeq™ 500 High Output Kit (150 cycles)
List Price (USD):$2,500.00

FC-404-1004 | NextSeq™ 500 High Output Kit (300 cycles)
List Price (USD):$4,000.00

FC-404-1005 | NextSeq™ 500 High Output Kit (75 cycles)
List Price (USD):$1,300.00

Based on this information I've worked out that the cost per library on the NextSeq is almost exactly the same as the cost on the Ion Proton with the upcoming PII chip, which I assume it was intended to compete with. https://docs.google.com/spreadsheet/...k8wczNWN0kxeFE

Please let me know if you have more information than I do.

[GenoMax]

From the NextSeq 500 system user guide from support site.

For any index sequences, RTA2.0 requires at least one base other than G in the first two cycles (i.e. i7 -705 index is not compatible)

[GW_OK]

Yeah, that'd mean an entirely new, machine-specific library kit and backwards compatibility issues.

Is this not just a little bit nit-picky, though? Think of a genome. Think of an amplicon. How many times would arise wherein the first X bases were a G-homopolymer? Theoretically it could happen, certainly, but practically, I'm thinking not so much.

Anyone have any concrete examples of G-heavy read starts?