We are an academic lab strapped of cash but need to do lots of qPCR in the coming months. We haven’t came across an undergrad that we trust with these assays so I am exploring the idea of automation. Not very high throughput (<100 96-well plates per month) but need walk away time so that I can do other stuff
I am eyeing mainly the Hamilton units right now because their tips are about half the price of what we use for manual pipettes (I was very surprised when my local Hamilton rep told me the price of their tips). Due to budgetary constraints I am considering 2 options that are both ~5k USD:
An open-box ML prep at a local surplus. Made in 2021, 2 independent channel configuration (no MPH)
An used Nimbus on ebay that seems to be reasonably maintained based on pictures. 8 independent channel configuration. Laptop and software not included.
I have sent both systems to my rep to see if they could support it. From my understanding the prep is supposed to be more user-serviceable while for Nimbus you need to pay someone to do it unless you run a farm of handlers. I also don’t know how much does Hamilton charge for the software. All things considered I expect the setup cost for the prep to be much lower.
On the other hand, I do feel the prep to be quite limited with only 8 deck slots and 2 channels and I should be able to do a lot more with Nimbus. It also appears that prep is very “locked-in” and there appears to be a larger community and ‘aftermarket’ for Nimbus.
So what do you think? Which route would you go if you are in my shoes? Is there anything I should look out for on these pre-owned handler? Haven’t used liquid handlers before but I did my both my college and PhD in engineering fields so I feel comfortable towards both systems. I am also open to other options that falls into the same price range.
I have purchased instruments from eBay before, but you need to fully vet what you’ll be getting. I’m very cautious with eBay stuff and it’s not my preferred route of getting used instruments.
The two other routes I’ve gone through before is Copia Scientific (they have anything that a lab would need, automation and otherwise) and Nugget Scientific. I would recommend either of these over eBay. And both of these companies also supply support for the instruments they sell. You can even purchase a service contract with the instrument to guaranty that you’ll get a well sorted instrument.
Prep and Nimbus will likely both run >$5000 - not including IQ/OQ and service to reestablish the system to working operation. PM costs are typically ~5-10% instrument cost (from mfr) yearly, so also consider this into your budget as well.
For an academic lab, you might look for an older generation Opentrons with a small volume pipette. These systems are good entry-level and are scripted in either Python or with an online GUI hosted on their website. Tons of pre-scripted methods are available and can be tweaked to fit your use case.
I don’t trust Opentrons pipetting as much as Hamilton, but OT honestly has a better usefullness:cost ratio than the big players like Hamilton/Tecan/Beckman; especially at the lower cost range.
Depending on your area, there are plenty of lab equipment auctions happening in the market recently. Check websites like EquipNet if you have time to burn finding a good deal in your area.
Yeah for 5k I was thinking only the initial acquisition. I know there is probably more to get it working but I couldn’t figure out how much.
For what Hamilton told me the prep is supposed to be setup by end users. They do sell an volume verification kit for “a few hundred dollars” but it sounded like that’s all what it takes to get it running (if there is no need for re-calibration/repair). The unit was still in factory box so I hope it has been okay on the shelf
You are right that the Nimbus will likely cost a lot more to setup which makes me hesitate despite how cool they are.
I actually started my search from Opentrons. But from what I gathered OT-2 requires constant calibration for both labware and pipette which I simply won’t have time for. It also don’t have liquid level detection and keep track of them a priori sounds like a real headache. The newer Flex supposedly solved all these issues but then the cost advantage is gone.
This is the tradeoff, of course. But there are plenty of labs that make do with a fleet of OT-2 in lieu of a monster STARplus or something similar. In most cases, once you validate a process once, its just tweaking and adjustment rather than full development. And even hand pipettes need annual or bi-annual calibration so that sort of maintenance isn’t out of the norm.
I think your final investment will likely be closer to $10k unless you find a steal, and you’ll likely need to find a balance between capability and price. For instance, even if you had liquid level detection - would you trust it? Or for that matter would you have the time to develop error recovery where LLD would even benefit your workflow, especially if it aborts protocol mid-way through MM or template distribution?
That being said, I personally love the ML Prep - get one if your budget accommodates. I think they are a great low(er) cost option that doesn’t sacrifice much from Hamilton’s larger scale capabilities.
I read that OT-2 calibration was needed every few days instead of every year. Some even say you need to do it before every run. Maybe things have gotten better?
Like I said I am currently leaning towards the prep because it looks like it is more likely to solve our current problems within the budget. It’s just my inside self still wants a bigger toy…
I think it really depends on what kind of labware you are accessing. If you are working with 384 well qPCR plates with very tight wells, then you’ll have to calibrate more often. We do qPCR with 384 well plates, and we have gotten away with quarterly re calibration if no hard crashes occur due to incorrect loading of the deck. If you only need to access 96 well plates with larger wells, then calibration may be required less.
For us its only used once per week though, so with your 100 plates/month that use that calibration might be required sooner.
If you were planning on using the Opentrons OT2 multichannels, then you will need to keep in mind 2 things. One is that the entire multichannel will always be the same volumes, so you lose that ability to pipette different volumes. This also means that if you need to do dilutions and they are variable, that you’ll need to use a single channel pipette and expect it to be slow, or . Second, if you use the multichannel pipettes at a high rate, you will need to inspect the O-rings carefully and change them out as soon as you see any damage to them. It’s very easy to start getting poor transfers or crashes because of this.
Having that said, the OT2 works great when it is taken care of. I agree with @evwolfson, you have to carefully consider this kind of tradeoff of price vs maintenance burden vs utility to decide what makes sense for you! There never is an optimal solution, its just what you can live with at the end of the day.
Update: the ML prep mentioned in OP was snatched by someone else so I am back to search for other listings. Also reconsidering OT-2. What questions would you ask the sellers? @evwolfson@jnecr
Service records and provider, service/maintenance schedule, instrument usage, serial number.
Call the Mfr with serial number to confirm instrument record. Call service provider to confirm service record, and research how caustic the chemistry being used is on the system. Alternatively, if the chemistry is extensively forgiving, it could be a sign that the system doesn’t need rigorous maintenance and upkeep to perform and you’d need more scrutiny on your end to ensure the system works well enough for your needs.
I haven’t played with OT instruments much so there might be some other red flags to look for that I don’t know. Above is just the basic work I’d do considering purchase of any core instrument. Maybe software and firmware versions or software dongles could be important.
@yl5325 This may be a bit of a longshot for you but have you considered checking with internal channels (surplus, etc.) at your academic institution? I am also at an academic institution and we have had the opportunity to acquire liquid handlers at a steep discount and in some cases free.
We have previously purchased 2 preowned Hamiltons. A STAR and a STARlet. The STARlet was from a trusted former colleague and went well. The STAR had 8 damaged channels due to aspirated liquid, which needed to be exchanged for a total of 20.000 € at the time.
Recently Hamilton is not as supportive anymore for pre-owned instruments. They require a hefty fee to get the instrument into the maintenance contract and limit upgrades to the software. Unfortunate but it is what it is.
That being said, your best option of the mentioned is probably the ML Prep. Easy to program, excellent accuracy and maintenance by the user. You may be able to upgrade with the 8-Channel tool as well, too. But check if it provides an advantage as these channels are connected.
If you need even more choices, may I suggest the PIRO? May not be common in the US but these German built robots are specifically made for PCR. They are workhorses, almost never break and can be programmed very easily. They aren’t fast by any stretch and are unfortunately discontinued. However, service is still available. PIRO
Have you considered pipette-free liquid dispensers, such as the Echo, Mantis/Tempest, iDot or CertusFlex? They are not necessarily cheaper but you don’t need tips and can downscale your qPCR reaction to 2ul.
Yeah, Hamilton told me they need $4.5k to just look at the unit, let alone any repairs. Service contract is >$8k/yr and you need 2 yr minimum. Very high CTO.
I looked at PIRO and they look very interesting. But unfortunately they apparently are not sold in US.
Good point, but we rarely see lab equipment in our university surplus, not sure why. And maybe these robots are more common on either coast? We are in midwest and I never see anybody using them except the big ones in core facilities.
Those are very cool technologies. But plate-to-plate transfer is not the bottleneck we are trying to automate here. Our thermocycler uses a 96-well block too.