Lynx for High Throughput Screening

Dynamic Devices (Lynx)
1

Hello! Does anybody have experience using the Lynx for high throughput screenings?

I’ve been considering getting a standard volume head and pipetting from compound libraries that are in 96 well PCR plates, serially diluting these compounds into various 96 deepwell plates, and then pipetting these dilutions into 384 well plates. I’m curious to see how people view the Lynx in terms of user-friendliness / flexibility / ease of programming

Cheers!

3 Likes
2

Also curious about this. Love the idea of 96 head with independent volumes, but worried about software support. Also not sure about it would handle hit-picking.

3 Likes
3

I was thinking of just using the standard volume version for my purposes, but yeah variable volume sounds great (my scientists don’t have a use for VV unfortunately)

5

They also have a 24 channel head which i heard is pretty slick when dealing with tubes

6

I’ve used them a bit so my experiences are a bit different perhaps than others.

They’re a smaller company BUT they’re expanding and rightfully so since they’ve cornered an in between market with some cheaper costs than the big liquid handlers and some slick, new tech (VVP).

With that said,

  • The software has some annoying bugs and they release updates almost weekly to try and address them (while also generating new bugs.)
  • The tips are not great and there have been supply chain issues. Although the new 200uL tip is much better I hear!
  • Static problems plague the tips. 35uL tips are a problem so install anti-static fans or inject them back into boxes with added tip shaking movement to ensure they’re actually off. They’re working on fixing this but it won’t help you right now to know they’re trying to fix it.
  • Documentation is among the worst in the business. Maybe this has changed in recent months!
  • Software decisions around the VVP are mind boggling (mandatory air gaps not being hard coded in the software) so demand proper training BEFORE doing any kind of development on the system or you will lose months in development time.
  • They changed the 96 head type and that improved a lot of the old problems so be careful if you’re buying used.
  • Software is easy and intuitive to use but in order to maximize usage, you really need better documentation with best practices.
  • Support can be iffy so get some training when they install.

Overall the Lynx is great if you’re doing basic tasks. It’s easy to write scripts for and the VVP tech is pretty neat. However liquid handlers require an ecosystem of support and it’s important to keep everything from roadmaps to engineering support to supply chain concerns in mind. Furthermore it’s important to keep your assays in mind. If you have an assay that is more quantitative than qualitative (and uL’s matter), I would not recommend them.

I know two companies with managers who dislike the Lynx, and at least one is looking to actively move away from them. With that said, there are probably many more places that love them!

2 Likes
8

Definitely, I know people that love them!

  1. I’ve never had a problem with Hamilton supplying tips but they pissed off a lot of people (including me) by just changing tip types without a heads up (outside of a marketing email) or by insisting that the hardware changes WERE going to happen whether you wanted them or not for the tip tech.
  2. I had no issues qualifying LC’s on an ARTEL but for volumes lower than 5uL, the final results were not as robust as on Hamilton’s or Tecan’s. The data was just not as consistent on the VVP. It could also have been a static problem or a lab environment problem that impacted the pipetting results.
  3. The benefit I’ve heard is that they treat large and small companies the same which is fine but the problems are always exponential for larger companies. If you’re processing millions of samples per years, the issues do not scale in a linear way.
3 Likes
9

TL; DR - Buy as many LYNX machines as you can fit in your lab.

The Dynamic Device Lynx are one of my favorite liquid handler. Regardless of your problem in lab. Development on them is orders of magnitude faster than other liquid handlers for once simple reason.

You can step thru a program.

Even more you can e-stop or unplug the thing and jump inside loops and resume.

So yeah tip offset loading, changing things into plate based, etc, it adds complications… but nothing you can’t overcome by utilizing the fast development environment it delivers.

Accuracy is good too. We used Artel on it and hit 3uL out of 300uL tips with ~ 3% cv. after some liquid handling parameter tweaking.

For cherry picking? Eh they have work to do on their independent axis arm, but it will get there.

The software yeah it looks like a windows 95 application, but it’s reminiscent of Biomek software UI. Don’t like it. Fine write your own software and control it using it’s remote service mode.

You can also insert in native C# scripts in methods so you aren’t constrained to writing batch scripts or compiled .exe programs. It’s nice having a general purpose language right there to get out of sticky situations.

4 Likes
10

What is the price range for a lynx liquid handler (< or > $50k?)

11

it depends on if you buy the VVP head or the SVP head. Both cases will be over 50k. I don’t know of a liquid handler under 50k besides the opentron.

Could also keep any eye out on the used market as well. @Dyanreagan from Copia has a good selection of used lab equipment.

1 Like
12

Thank you!

13

When I asked at SLAS they gave me a rough estimate of around 100k USD for a Standard Volume Pipette with around 18 slots (I believe that’s public knowledge). It sounds pretty reasonable to me at least

2 Likes
14

Hi! Nice to meet you - is there anything we can help you with?

15

Not currently, but thank you!

16

YW!

18

Do you have the 8 channel on a Lynx?

20

I don’t know if they changed it but it was a big time disaster at launch. A year came and went and it was still a disaster.

New tech is fun and cool but early adopters risk it all. I’ve seen people lose jobs, tarnish their reputations and change careers because they trusted sales guys or thought that because Vendor does A well then B should also be good. Always proceed with caution.

4 Likes
22

I think depending on your line of work, “just fine” is not good enough.

One of the one’s we had was a dud and the other required a lot of a babysitting which I believe some
Would qualify as “just fine” but not worth the time or savings in the long run imo. Should have just went with something more established and have saved ourselves on the costs of engineers/downtime issues.

1 Like
24

What exactly were the issues? I’ve always found downtime due to maintenance/ calibration is the biggest bottleneck for production robots. Especially if it’s something only a service technician can fix. This can be a deal breaker even for robots that have a lot of other good qualities.

I’d prefer if people always took each other’s posts in good faith :slight_smile: we are sharing expertise around major capital purchases and operational expenses, people have a ton of different opinions.

2 Likes
25

You nailed it. In high throughput, large volume environments those bottlenecks get amplified by 10x-100x. If I were operating a small volume lab or just R&D, I would have a different set of experiences. However my experiences are on a scale of multiple millions per year. 24/7, 365 labs.

At scale, you end up valuing support/service way more than design aesthetics or new tech. That’s why it’s super hard to breakaway from the time tested players in the field.

The service side was a problem but also the manufacturing of the 8 channel was a big bottle neck. I believe the company has addressed those bottle necks now. But I can’t speak to the 8 channel since there were other issues related to manufacturing as well which made the service complicated at first.

Edit: also they train people now on how to do repairs for it but that wasn’t available up front. With the 96 head, there’s way more you can do on your own tho.

1 Like
27

Definitely not! Just wanted to share my experiences.

Wish we had something like consumer reports for lab automation stuff!

2 Likes