Started our first HiSeq2500 run.

[pmiguel]

Doing the 2x150, of course. (BTW, a little cumbersome -- had to buy three 200 cycle kits, split one between the other two.)

Okay, trying to think what would be salient for those awaiting or considering upgrades.

We did the on instrument clustering. That does save a lot of time. Just plink your flowcells in place (same process, except no cBot involvement) and put your denatured libraries in open 1.5 ml tubes in the two slots provided.

Much smaller volume of cleave for this chemistry, which means it is not the bottle that is the last to thaw.

Washes while in rapid mode are also rapid. A maintenance wash takes less than 1 hour.

There are fewer reagent bottles, so you end up needing to provide some 250 ml bottles of your own to put 'PW1' into, just to keep the ends of the unused ports wet.

The new syringe pumps are fast! Had some issues with them during install though. One bad syringe (was delivering 1/2 the correct volume) and one bum solenoid. For those of you who lived through the days of the bad v2 HiSeq/HiscanSQ solenoids, you may have felt a chill reading that. Hopefully an isolated occurrence.

Waste lines: those that terminate what would be lanes 1,2,3,6,7,8 for each flow cell now do nothing. However, to prevent them from clogging up while sitting unused, you put them in a beaker of fresh water. That water gets pumped in and out during the run.


Going back to standard run mode -- requires a 4 hour wash! Obligatory! So don't blithely decide to do a standard run and then change your mind.

I don't know if it is possible to do rapid chemistry in one flow cell and standard in the other. I think I would try to avoid this, even if possible.

I will do another post after the run finishes (Wednesday? I guess. We are on cycle 108. Run started about 6 PM yesterday. Also specs call for the run to take 40 hours.) Wow, that is fast...

--
Phillip

[pmiguel]

I forgot to mention, the turn-around chemistry is placed on the instrument when you start the run -- so no further interaction required.

--
Phillip

[pmiguel]

The Good Flowcell

One of the flow cells had some problems with focus for some of the swaths. Not sure why -- not the usual bubble in the low bottom tile issue.

Here are the results from the other flow cell in the form of a few graphs. If you want to see some other results, let me know what and I will post it.









--
Phillip

[luc]

Thanks Philip,
results seem to look great!

[pmiguel]

The not-so-good flow cell.

These were DNA libraries. The other flow cell comprised RNA libraries.







Here you can see some tiles had horrendous error rates:


These were entire swaths that were mis-focused. Illumina tech support attributed them to the usual BMS 'bubble in focus tile causes instrument to mis-focus'. I don't think so because I did not see a bubble in these tiles. But tech support thinks they might not be visible due to the mis-focus.

Anyway, the %Q30 puts the run well within specifications for 150 nt reads.