Buying GAIIx now?

[vebaev]

Hi all,
My department is buying a GAIIx and I was wandering is there a chanse after few, 2-3 years to become not supported or too old fasioned to work with (since increasing the read lenghts of other new models)? But after all we have a budget for that and now, so it will be this or nothing?

What do you think?

[ETHANol]

My guess is in a couple years it will be too expensive to use, i.e. the price of sequencing will be much cheaper with newer machines. No one will want to run their samples on it because they can pay someone somewhere else to do it for less. You also have to ask the question of are there the funds to run it. Is there the demand for it at your institution because it is very unlikely that people outside it will go to you. You will need to fill all eight lanes with every run that's a lot of money in consumables and from the sounds of it, money isn't exactly growing on trees in Bulgaria. Usually, it is easier to get the funds to buy big ticket items then it is to get the funds to use them. What is the per base/read cost? And how does that compare to sending your samples outside to a core facility. How much more sequencing could you do if you used the funds to send your samples out somewhere? Are the funds 100% dedicated to buying a sequencer?

I know there is the desire to bring cutting edge genomics to (Insert your small country/institution here). It also looks good on grants to have all sorts of fancy equipment at your institution. But on a pure science level, I bet it sits there collecting dust because no one can afford the consumables to run it. Balkan science at its best.

Is the goal to bring cutting edge technology to Bulgaria that makes your institution look great or is it to do cutting edge science and publish in high impact journals? If your goal is the first then buy it, if it is the second then probably don't.

That's my take from the Balkan perspective.

[vebaev]

Yes I know what you are tolking about (I was 2 years in Greece). But nevertheless the funds are dedicated for equipment and we cannot relocated it for services or materials for other thing..

[Heisman]

Could you maybe purchase a MiSeq (or is it MySeq) instead?

[vebaev]

With MiSeq is quite limited to smallRNAs and target sequencing, you cannot do whole transcriptomes or genomes...

[ETHANol]

Not to mention cost/MB or cost/read of the MiSeq. With the GAIIx you are already paying a premium price for every read or MB sequenced. Add on the extra cost associated with the inflated price I'm sure you pay for consumables in Bulgaria (How Balkan pricing works: Balkan price > German price > US price > discounted price for heavy users).

Another thing to think about is where is sequencing going. There are two possibilities (or the combination of buth) I can think of: 1) It goes the way of oligonucleotide synthesis and capillary sequencing, i.e. it is moved from academic core facilities to highly efficient for profit companies like IDT or Macrogen. 2) It becomes like the PCR machine and every lab has a couple of them sitting on their benches.

Whatever it is, in two years things will be a lot different. If you are not running the machine full time, it will not be worth the money and even at this point it may already be past it's usefulness.

http://onlinelibrary.wiley.com/doi/10.1111/j.1755-0998.2011.03024.x/pdf

[james hadfield]

I would reccomend starting with a good service provider and getting the experience in sample prep and bioinformatics. I think an excelent informatics team could be assembled for the cost of a seqeuncer and genomes are getting cheaper and cheaper to buy. At the end of the day the sequencing is much easier than the analysis but I guess it does lok god to have the shiny new toy in the lab.
You can buy refurbished GAs however the cost of dta generation is much higher as you wil only get 40M reads per flowcell compared to 200-300 on HiSeq.
You can find a list of core facilities on the SEQanswers wiki or on the Gooogle map. The cost will depend on your experiment but 50 genomes or more and this starts to plummet.

[james hadfield]

I would reccomend starting with a good service provider and getting the experience in sample prep and bioinformatics. I think an excelent informatics team could be assembled for the cost of a seqeuncer and genomes are getting cheaper and cheaper to buy. At the end of the day the sequencing is much easier than the analysis but I guess it does look good to have the shiny new toy in the lab.
You can buy refurbished GAs however the cost of dta generation is much higher as you wil only get 40M reads per flowcell compared to 200-300 on HiSeq.
You can find a list of core facilities on the SEQanswers wiki or on the Gooogle map. The cost will depend on your experiment but 50 genomes or more and this starts to plummet.

[pmiguel]

You can buy refurbished GAs however the cost of dta generation is much higher as you wil only get 40M reads per flowcell compared to 200-300 on HiSeq.

Not sure what the typical yield is on a GA, but a HiSeq would produce 150-200 M reads per lane using v3 SR chemistry.

--
Phillip

[GenoMax]

You have said elsewhere that the funds can't be re-purposed for services but hopefully you will be able to convince the folks in charge otherwise after seeing some of the comments here.

If you can't then you should look into possibility of securing a long(ish) term contract for reagents at a reasonable price (and a guarantee for availability). If you are unable to get one or both then that could be used as a red flag for the local "powers" to reconsider this purchase. I hope you have a separate budget for purchasing the consumables/reagents besides the hardware.

Cost of the associated informatics is altogether a separate consideration.

Hi all,
My department is buying a GAIIx and I was wandering is there a chanse after few, 2-3 years to become not supported or too old fasioned to work with (since increasing the read lenghts of other new models)? But after all we have a budget for that and now, so it will be this or nothing?

What do you think?

[protist]

For the GAIIx with v5 kits and SCS2.9/RTA1.9 you can get a yield of 40-50 million reads per lane (for a 42 cycle run).

[xristi]

Hi,

I'd recomend buying HiSeq 1000 instead of GA2x. It is newer technology and upgrade/support path is very similar to the HiSeq 2000 which is widely used in the field.
GA2x has been around for a while but it's much older instrument than the HiSeq product line.

[gogreen]

I also would recommend against investing on a GAIIx at this point. The Hiseq has been around for a couple of years already, so the GAIIs sound ancient in terms of the data generated and costs incurred on the HiSeq v3 chemistry. As xristi said, a HiSeq 1000 would be a better option or one can always use the service facilities or collaborate with academic labs with the infrastructure.