Header Leaderboard Ad

Collapse

Cost/run on HiSeq 2500

Collapse

Announcement

Collapse
No announcement yet.
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Cost/run on HiSeq 2500

    Does anybody have an estimate of the cost per run on a HiSeq2500?

  • #2
    HS upgrade

    We were told 2500 cluster kit for PE is $1225; SR is $795. 200 cycle SBS is $1690, 50 cycle SBS is $465.

    Has anyone yet upgraded and how is it working out?

    Comment


    • #3
      Well that would be great.
      But that would be a per lane reagent cost below that for a lane of HiSeq 2000 reagents. Since the amount of sequence generated per unit time is similar for a 2500 vs a 2000 flowcell, I will speculate that the person feeding you this information is wrong.
      But maybe Illumina is just planning to steal the wind from the Proton Torrent's sails. This would tend to have that effect.

      --
      Phillip

      Comment


      • #4
        Originally posted by pmiguel View Post
        Well that would be great.
        But that would be a per lane reagent cost below that for a lane of HiSeq 2000 reagents. Since the amount of sequence generated per unit time is similar for a 2500 vs a 2000 flowcell, I will speculate that the person feeding you this information is wrong.
        But maybe Illumina is just planning to steal the wind from the Proton Torrent's sails. This would tend to have that effect.

        --
        Phillip
        These are the costs for the '2' lane (single sample) flow cell, in rapid mode run--39 hrs for 2x150. MiSeq equivalent. By our calculations, it costs more, but makes operation more flexible--hard on users who want short, single reads to have to go in on 2x100 nt run and be charged accordingly.

        Hilary
        Attached Files

        Comment


        • #5
          Our rep indicated a premium of ~15% per GB data vs. the HiSeq2000. That's exactly in line w/ HMorrison's prices, since it would take five runs (120GB per HiSeq2500) to equal a single HiSeq2000 (600 GB).

          Comment


          • #6
            Apparently I miscalculated.
            Still I would pay 15% more to get a data set back in 27 hours, rather than 12 days.

            Hilary, what does the "same sample" comment in your spreadsheet mean? Are the 2 lanes in the flow cell loaded from a single source?

            --
            Phillip

            Comment


            • #7
              Hi all I think the clustering is going to need to make use of current fluidics. It would be surprising if the $50,000 cost included new fluidics. If this is the case then I am not sure there is space to squeeze in all the reagents necessary to perform different run types. I guess we'll have to wait and see.

              It would not be a big hardship to have to run the same run on both sides. The upgrade could be what many core facilities are looking for to get additional flexibility.

              I posted a bit about the Proton install at BCM and HiSeq 2500 rapid runs on my blog...http://core-genomics.blogspot.co.uk/...me-in-day.html

              Comment


              • #8
                Hilary, what does the "same sample" comment in your spreadsheet mean? Are the 2 lanes in the flow cell loaded from a single source?
                Yes, from what Illumina's told us earlier this year, clustering will be performed on the Hiseq2500 and the same sample (or pool of samples) will be sequenced on both lanes.
                I guess the sample (or pool of samples) will be loaded on the paired-end rack at the beginning of the run.

                Comment


                • #9
                  This is exacty what I have heard as well. The same pol of samples will need to be rn on both lanes of the 2 lane flowcell. Has anyone seen a flowcell yet? I mocked on up for my blog last year, but have not seen a real one yet.

                  Comment


                  • #10
                    Originally posted by pmiguel View Post
                    Apparently I miscalculated.
                    Still I would pay 15% more to get a data set back in 27 hours, rather than 12 days.

                    Hilary, what does the "same sample" comment in your spreadsheet mean? Are the 2 lanes in the flow cell loaded from a single source?

                    --
                    Phillip
                    Yes, although there are two 'lanes' they have to be the same sample--I think loaded from the same port. Sorry for the delayed response. Have been fighting with the FLX (and about to give it up).

                    Comment


                    • #11
                      HiSeq 2500 units delivered

                      Does anyone know how many hiseq2500 units have been delivered?
                      Has BGI upgraded their 127 hiSeq 2000s? Anyone?

                      Comment


                      • #12
                        We just got our machine upgraded to a 2500. Validation runs went great and we will be running two FCs today on Rapid Mode! Nice not having to wait to fill FCs when odd read lengths are requested. There is data available on the Illumina BaseSpace for anyone to take a look.

                        Cost is as below (list prices):
                        SBS (50 cyc.) = $465
                        CLuster kit for on-board clustering = $795
                        cBOT clustering (additional) = $400

                        So for a 50bp (2 lane) rapid run ~300 million reads
                        If done on board = $1165
                        If cBOT clustering done = $1660
                        Per lane = $830

                        For 100bp paired-end, kit prices are as blow:
                        SBS (200 cyc.) = $1690
                        CLuster kit for on-board clustering = $1225
                        cBOT clustering (additional) = $400

                        Comment


                        • #13
                          Originally posted by HMorrison View Post
                          These are the costs for the '2' lane (single sample) flow cell, in rapid mode run--39 hrs for 2x150. MiSeq equivalent. By our calculations, it costs more, but makes operation more flexible--hard on users who want short, single reads to have to go in on 2x100 nt run and be charged accordingly.

                          Hilary
                          How to calculate the per run cost? Thanks

                          Comment


                          • #14
                            After looking through this thread I am still at loss as to what is the cost of the entire run on 2500 or 2000. My problem is that my libraries have to be sequenced with my custom sequencing primers, that is my samples cannot be pooled with other samples that use standard sequencing primers. The libraries run fine on MiSeq (2x100 PE) with these custom sequencing primers, but I need a much higher output than a few gigs per run. Thus, for 2500 I would nee to buy a two lane run, but on 2000 I would have to buy an 8-lane run, correct? So what average cost I should expect from a commercial provider (I need only fastq datasets)?

                            Comment


                            • #15
                              Originally posted by yaximik View Post
                              So what average cost I should expect from a commercial provider (I need only fastq datasets)?
                              Here is an example cost estimator you can try (without having to contact anyone): https://dugsim.net/estimate_cost

                              Following sites give you access to multiple providers for quotes:

                              AllSeq: http://www.allseq.com/
                              Genohub: https://genohub.com/
                              Last edited by GenoMax; 01-08-2014, 11:18 AM.

                              Comment

                              Working...
                              X