Brown University



We are pleased to announce that the Genomics Core Facility has upgraded its Next Generation Sequencing Operation.


The Illumina HiSeq 2000 sequencer has been outfitted with new hardware and software to allow operation in two different modes: A high data output mode (HS2000 mode) and a fast throughput mode (HS2500 mode). The HS2000 mode is the same as we have been running up until now, so it is the add-on HS2500 mode that provides some additional capabilities.

The HS2500 mode uses a different flowcell, with a 2-lane configuration, as compared to the 8-lane flowcell that continues to be used for the HS2000 mode. Instead of dedicating one lane of a 2-lane flowcell to a control, Illumina recommends using spike-ins of PhiX.

TruSeq compatible libraries which were made for sequencing in the HS2000 mode are fully compatible with sequencing in the HS2500 mode.

The two modes cannot be run at the same time on the instrument (i.e. it is not possible to use a HS2000 and a HS2500 flowcell at the same time).

The HS2500 mode uses a new clustering procedure that is performed right on the HiSeq sequencer, eliminating the need for the cBot. HS2000 flow cells will continue to be clustered on the cBot. There is one limitation to clustering in HS2500 mode, namely, the same sample has to be run in both lanes. In most instances this should not present a problem, as it can be compensated by adjusting the multiplexing schemes. For example, if you wish to sequence 6 libraries, and you think you need 2 lanes to get enough reads, in the HS2000 mode you would multiplex 3 libraries per lane, and run 2 lanes, the first with libraries 1-3, and the second with libraries 4-6. In the HS2500 mode you would multiplex all 6 libraries in 1 sample, and then run 2 lanes with that sample.

If a scheme like this is not possible, different samples can be run in the 2 lanes of a HS2500 flow cell. However, in this case clustering has to be done on the cBot, which requires additional reagents, and hence will incur an additional charge.

In terms of the amount of data you can expect in the different modes, Illumina claims 180-200 million reads per lane run in HS2000 mode. In our experience, the "real life" yield falls in the range of 160-180 million reads (approximately a 10% deficit). For the HS2500 mode, Illumina claims 130-150 million reads per lane. As we do not have any hands on experience with this mode, we do not know whether the same 10% "real life" deficit should be applied (but it is probably reasonable to expect that there will be some shortfall). So you can expect ~75% of the data (reads) in the HS2500 mode that you would get in the HS2000 mode. The compensating factor is time, as the runs in HS2500 mode are much faster. An additional important issue is queue time, since as soon as we have 2 samples to fill a HS2500 flowcell, we can start a run. So the wait times to get on the instrument should also decrease significantly.

  • HS2000: 100 bp paired end, ~11 days
  • HS2500: 100 bp paired end, 2-3 days
  • HS2000: 50 bp single read, ~3 days
  • HS2500: 50 bp single read, 1 day

The pricing for the HS2500 mode will also be different. We are working with the Controller's office right now and will announce the new prices in a few weeks.

We anticipate that the HS2500 mode should be available for users in a few weeks. It has been installed and Illumina has completed all their QC procedures. We are in the process of running an in-house "real life" sample, and if that goes well we should be ready for business.