Does anyone have any experience with using imitation liquids to simulate expensive or low-availability reagents for liquid class development? For example, using artificial blood in place of real blood. I’m curious what people’s experiences are on what is and isn’t effective. Are there any “recipes” for good imitation liquids? Is there anything commercially available (that doesn’t break the bank)?
Applications of highest interest to me would be lysates and PCR master mixes.
What parameters are you most interested in, viscosity? If so, you could always mix water and corn syrup or glycerol for various degrees of this property. Just be certain to not overaspirate, of course 
.
While it may be on the pricier side, Artel offers liquid class development with PCR Master Mix-esque solutions that can be verified volumetrically through their system. https://www.aicompanies.com/artel-liquid-handling/mvs-qualassure-solutions/pcrmix-qualassure/
I’m unsure how to best replicate lysates myself, hoping this spurs some good discussion on this topic!
I wonder if Genie Life Sciences has any insight into this. Their LC library takes density, viscosity, and other properties as inputs and provides pipetting techniques based off those parameters rather than by liquid name.
Surely they used some sort of analogues to optimize this? Just not sure who to ping for visibility.
This is the future and also where AI can have one of the greatest impactful uses!
The Tecan Fluent has a similar tech where you feed the LC a density and viscosity variable and a pre-trained neural net handles the rest. It also is tip agnostic so those two inputs apply for for all tips, no need to set it up again when you change tips.
Hi Cole,
What I’ve found is that if you’re testing for liquid class volume handling accuracy, you want to look at viscosity, polarity, whether it is a detergent, and sample-specific affects.
Viscosity
For most “relatively polar” liquids, which are essentially just water based liquids, you can mimic general dynamics on a liquid handler by finding the most comparable % water/glycerol solution to solve for viscosity.
Polarity
Polarity is a bit more tricky, especially if you’re testing a non-viscous solution. The closest material is an alcohol mix, which is less polar than water but extremely non-viscous. This often becomes a problem for when liquid handlers need to keep the material in the tips over a relatively longer period of time, as many of the forces such as friction and polarity that help retain the vacuum of the liquid in its tips are diminished.
Detergents
If you’re trying to get a detergent class liquid (such as tween) to work in a liquid handler, just stop. Even at its most diluted, accurate liquid handling will be a nightmare from the bubble formations alone. If you really must develop a detergent containing liquid, use A DASH of tween in your solution, but understand that it will never truly leave any surface it ends up adhering to on you instrument.
Sample specific considerations
The best course of action to find sample specific type mimicry liquids is by looking outside the lab environment. Depending on the application, there are several industries that require this type of material that are not necessarily lab related. To answer your example with blood, I found this amazing paper that discussed the issue in how to accurately mimic the non-newtonian properties of blood to be used in their 3D printed medical simulators for things like surgery planning. Most industries that currently used fake blood in things like Ultrasound tech training use a 60%/40% water-glycerol solution but it was not sufficient for their purposes. They created a home brew recipe that followed the same fluid dynamics as blood through the course of their research and experimentation.
Radware Bot Manager Captcha
I used this as a pet project last year and it acted relatively similarly on my 8-span head, but the differences in performance were so small in regards to viscosity that it didn’t make a discernable difference in the tests I was running, so I didn’t pursue the matter further. It also didn’t address my main problem which is blood’s coagulation due to exposure of air. Nevertheless, it was a phenominal paper that truly helped me understand the dynamics of blood as a liquid class, and taught me how I might be able to find solutions to liquid class problems in the future.
In a different scenario, I wasn’t able to get a Yeast culture protocol to work correctly in production because the cells in the sample were so adhesive to each other that changed the fluid dynamics of the tips they were being aspirated into. I used their growth media as my test liquid class, but it still wasn’t sufficient. I ended up having to use excess cells from other projects and my own DEV yeast culture to accurately mimic the dynamics of that liquid. Sometimes, you just got to get the real thing.
For our course of work, most liquids can be mimicked by creating combinations of polar/non-polar and viscous/non-viscous solutions that most likely resemble the liquid class in question. For PCR mixes, my understanding is that they act pretty much like water, unless they have some special property that changes their viscosity. For Lysates, I’d go through the process listed above, with special consideration on how the lysate cell debris interacts with itself.
At the end of the day, there is no avoiding using the actual sample material to validate your process before being put into production, so duplicating the exact fluid dynamics for each assay developed isn’t really necessary, and the above guidance should be sufficient. I’ve had great success with it in the past.
Let me know how it goes, and good luck!
Best,
Doug
Unfortunately, despite their array of liquid classes you can choose from on their devices, each one only relates back to 1 of 4 base liquid class definitions it uses in their software. For instance, somethings like “blood plasma” and “BSA” would both just be considered their “Serum” base liquid class. Most liquid handlers these days come with those base classifications, and I’m not a fan how Genie implies that it has defined these liquid classes more rigorously than other companies when they use the same base liquid class definition system.
They had me fooled. That’s good to know though. I guess there’s no silver bullet for liquid classes outside of validation and optimization