I am developing a method that will - essentially - transfer plasma from an input tube to a plate. The input tube is a 5mL Nunc 2d barcoded tube.
I had to develop the labware myself, but I took the measurements from the supplier and measured it with a caliper. I feel confident that it is as correct as possible.
Now, during transfer I have to transfer a volume of 4000 uL (Or anything between 3000 and 4000) so I divide the volumes in batches of 800. This works great. However it sometimes (mostly on aspiration 5) happens that the sample is detected as insufficient liquid.
Back in the day I used a recovery option of “available” but since this makes the data handling a bit complicated, I decided to just “repeat” the action and (if it fails again) exclude the channel. This works great.
However, my recovery here is to visit all samples that are marked as “Insufficient liquid errors” again and try to aspirate again something from them. This because I need to get the most out of the containers. This also works quite nice, my samples are retried again.
Now comes the issue.
I understand that insufficient liquid originates from the definition. The LH expects to find a liquid at X mm height but doesn’t find it there and thus prompts insufficient liquid error. But in theory my samples should be fine.
My question in this:
How to I minimize the “insufficient liquid error” by setting the labware definitions
I am curious to your errorhandling in this. For me it is important that I obtain as much as possible of the sample, but I also need to have the data of the transferred sample (so if there is still 800 left and we aspirate 800, I need that number for my data handling)
Can you clarify a couple aspects of your application? When you say you need to obtain as much as possible of the sample while requiring “data,” do you mean you want a measure of the exact volume transferred? For instance, if you had 3600 µL of plasma present, you would aspirate 800 µL four times, then the remaining 400 µL. Do you want the instrument to detect that only 400 µL remain, and to report that appropriately?
Additionally, are these samples relatively pure plasma, or do they have other components in them that could impact the liquid level detection?
As for the error you’re seeing, you may want to look more into how your labware is defined. Maybe your dimensions for the container are as correct as possible, but the position of that container in the tube carrier is higher in Z than you expected. There are also some bugs in Hamilton’s labware editor in the calculation of liquid volume within a container based on height–unlikely to be a significant issue here, but worth looking into.
This error also depends on the “submerge depth” that you specify. If all of your labware definitions are correct, then I would suggest attempting to minimize your submerge depth and make sure liquid following is enabled on aspirate (if you haven’t tried this already).
Let me know if any of this is helpful! I’m interested to hear what others think.
Basically I receive a worklist with a provided volume and I need to take as much of that volume as possible with a maximum of 4000 with 800 ul per loop. If I have 3760 as a volume, my last loop would have 560 ul as volume (3299 - 3760)
The sames are plasma, I do not see any indications of other components.
I am aware of the bugs in the labware editor. But it’s difficult what needs to change. However I have never investigated this option, so I’ll look into that more specifically.
Submerge depth is 2mm, lowered to 1mm but still in error. So for now I have used the errorhandling to be “retry” 2x and then "available’ since this will “disable” MAD.
Insufficient Liquid Level error is calculated based on a few factors:
Container measurements (segments)
Coordinates of the bottom of the well (Teaching)
Submerge depth
Volume being aspirated
Height in which the aspiration starts (LL - Submerge)
The container measurements are used to calculate how far the channel must go to aspirate the specified volume. The channel makes this calculation after finding the liquid level and compares this distance needed from the Liquid Level - submerge depth. If the final Z is below the bottom of the well, then Insufficient liquid level error is triggered.
If you trust the container measurements, I suggest checking the teaching of the source positions. If your Z is shifted higher than actual, you are limiting the liquid following distance and will trigger more aspiration errors. I personally teach labware to the absolute bottom with a very slight touch and adjust the dead volume measurement in the container. This won’t affect your pipetting like a relative shift of the labware does. Liquid following will continue down till this height and then stops.
If this continues to be an issue, you can try using MAD/TADM to stop pipetting when air is aspirated followed by a drain tip in jet mode for your dispense. You can use the following library to query how much volume is in the channels:
For insufficient LLD errors, you can recover from the bottom and still trigger MAD/TADM errors if air is aspirated. It can’t tell the difference between air that was aspirated to trigger the MAD/TADM error and sample, but it will be in the ballpark. You can use these values to update your sample tracking afterwards with the total volume that is in the channel.
The definition is something is created based upon measurements taken from 3 tubes. I feel these measurements are correct, but I will do some other measurements today and adjust if needed, same as the checking of the teaching.
Important to say here is that I am using widebore tips (3.2mm) but since I use the official library for that, I presume this is no influence to the cLLD.
Submerge depth was set to 1mm in my pipetting smt in general. Volume in the well is measured (by a manual pipet) to be exactly 4200 uL, the fourth aspiration of 800 works correct, but the last one (In theory there should be 1000 left) often triggers an insufficient liquid error. I changed the errorhandling to have 2x repeat and then available.
I’ll be in the lab in an hour and report my observations from this asap.
I have checked the labware definition and took a vial and did the measurement all over. There was a slight issue with some parameters but now the overal volume is not 4,2ml but 5,3ml. Which is a bit higher, but should be fine here. Teaching was done with 0.4mm lower. It ain’t much…
Hi @BrandonBare_Hamilton , unfortunately, the link to the QueryAspiratedVolumes library is no longer accessible. Would it be possible to upload it again?
There are two things that you should consider,
1 death volume or minimum volume that can be detected or aspirated for each type of containers. For the normal aspiration, there should be some volume which can not be transfered.
2 aspiration parameter and mode. You can test the difference between aspirate and aspirate all.
In your problem of insufficient liquid error, the instrument detected liquid, but the volume between bottom and submerge depth is not enough by calculation (volume of submerge depth was not included). So I think that you can change the aspirate mode from aspirate to aspirate all, then volume calculation will be omitted and the channel will aspirate all the liquid unitil stop at the bottom.
But in this way, there is a risk that the volume of aspirate is not right for existence of death volume. You can try TADM if the volume is fixed.
You are correct! In our workflow this is a bit problematic because we work with an unknown liquid volume. Ideally I transfer 4mL from a NUNC tube to a Axygen 10mL DWP. But that 4mL could also be less.
To overcome aspiration issues with those samples, I obtain the cLLD return and calculate the estimated amount of liquid. If this is ~10% more than my requested volume, I correct the aspiration volume accordingly. Same applies when there is lower volume. I simply recalculate to aspirate 60% of the volume “measured” by cLLD returns. This works for us, because our samples could have 4mL, but also 2mL and one huuuuuuuuge clot.
Aspirate all is not prefered. But this is more due to somebody in our QA department (hello QA!) that finds that “aspirate all” sounds scary.
