Chromatography Forum

Hi all.

We are doing a project on UPLC, and we have a "simple" question.

We are to calculate a final concentration from a measurement on UPLC as follows.

each sample:

5ul sample diluted in 50 ul buffer (1:10 dilution)

5ul 1:10 dilution in 100 ul precipitation solvent (1:20 dilution)

vortex + -20 degree precipitation for 15 min + spin 3000g for 10 min 4 degree

100 ul supernatant transfer to new tube and evaporate to dryness

Reconstitute with 25ul running solvent and load into UPLC

The concentrations are measured from a std. curve of pure anhydrous powder dissolved in running solvent and a 2-fold dilution series are made (concentration in uM)

However, the std samples are not prepped as the samples.

We hereafter interpolate "unknowns" from the std curve using graphed prism

We now just need to know if these steps are correct:

The final values for the samples are multiplied first by 10 and then 20 to get the final concentration in uM, as the std. curve ??

Thanks

"5ul sample diluted in 50 ul buffer (1:10 dilution)"

What volume do you get when you add 5 ul to 50 ul, and what fraction of that volume came from the sample ?

Hint; the answers are not 50 and 1/10.

Similarly with the next step and some lower down.

Peter

5ul out of 55ul (50ul + 5ul) = 11 fold

5ul out of 105 (100ul + 5ul) =21 fold

5ul out of 55ul (50ul + 5ul) = 11 fold

5ul out of 105 (100ul + 5ul) =21 fold

Which gives you answer to: "We now just need to know if these steps are correct:

The final values for the samples are multiplied first by 10 and then 20 to get the final concentration in uM, as the std. curve ??"

Peter

This is why I despair. You can be pretty sure that anyone writing a method where they say "5ul sample diluted in 50 ul buffer (1:10 dilution)" actually intended to make 5uL up to 50uL and thereby do a 10-fold dilution. Unfortunately you cannot be sure whether someone claiming to have followed the recipe did what the first part of the recipe says (added 5uL to 50uL to end up with an 11-fold dilution), or what they thought the recipe-writer intended (from the second part of the recipe), and diluted to 50uL.

So before we've even approached an instrument, we've already got an uncertainty of nearly 10%.

This is the sort of (very common) issue that totally messes up accurate work. And yet it is discussed far less than the relative merits of weighted versus unweighted fitting of calibration curves, which is a piddling detail by comparison.

The complete purist would, in any case, want to check that the diluting solvent and sample solvent are sufficiently similar that there isn't a change in volume on mixing.

This is why I despair. You can be pretty sure that anyone writing a method where they say "5ul sample diluted in 50 ul buffer (1:10 dilution)" actually intended to make 5uL up to 50uL and thereby do a 10-fold dilution. Unfortunately you cannot be sure whether someone claiming to have followed the recipe did what the first part of the recipe says (added 5uL to 50uL to end up with an 11-fold dilution), or what they thought the recipe-writer intended (from the second part of the recipe), and diluted to 50uL.

So before we've even approached an instrument, we've already got an uncertainty of nearly 10%.

This is the sort of (very common) issue that totally messes up accurate work. And yet it is discussed far less than the relative merits of weighted versus unweighted fitting of calibration curves, which is a piddling detail by comparison.

The complete purist would, in any case, want to check that the diluting solvent and sample solvent are sufficiently similar that there isn't a change in volume on mixing.

Amen to all that.

If it was ml as the units I would just go with the dilute and make to volume route. But how accurately can anyone make up a volume to 50 ul ?, and what do they measure it with ?

Peter