The HiSeq X Flowcell

Summary: Possibly 30% of reagents are wasted in dead volume on the HiSeq X. If you believe that reagents costs dominate in sequencing costs this could be significant.

Today I’ve been taking a quick look at the HiSeq X flowcell. The empty flowcell weighs 5.996g¹ . Fully filled with water it weighs 6.122g, which suggests that each lane contains ~15uL of fluid. I tried pipetting 15uL into a lane and this seems approximately correct.

I also disassembled a flowcell. The stack up is shown above. The coupling channel/hole seems to be made through the thick glass substrate². The flowcell channel (where sequencing occurs) height seems to be ~110 microns.

As channels are ~2.5mm wide and 64mm long this would suggest a volume of 17.6 uL. Overall channel volume seems to be in the 15 to 20 uL range. Based on the previous measurements this suggests that dead volume (reagents sitting in tubes) per cycle is ~50% of the flowcell volume.

Personally this doesn’t feel like a massive problem as I suspect reagent costs are a fraction of the cost of the total run cost. But if you believe that sequencing cost correlates with reagent usage here’s a case there Illumina are literally throwing money down the drain.

1. There seems to be some slight variation is weight between flowcells here.

2. The substrate seems to be exactly the same size and thickness as a regular microscope slide with a corner taken out.