Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

While reading this question: Why do we still not have an exact definition for a kilogram? , I had a crazy thought.

Using PCR, you make a known number of copies of a DNA strand where the length and composition is exactly known for all the copies. The number of copies is two to the power of the number of PCR cycles, and if you start with, say, all C-G DNA, you get a known number of duplicates which are the same molecular weight. If you only provide C and G nucleotides, you don't even have to worry too much about errors in copying, because whichever way they are fixed to make pair-matching DNA, you will get the same atomic weight. You can separate out the DNA and get a known huge macroscopic batch size of macro-molecules where we know the atomic composition exactly.

So we can easily make a batch of atoms where we know the number of atoms with potentially exact precision. Can you use this to make a mass standard? Just weighing the DNA won't work, because the water content will be uncertain. Drying the DNA won't work, because there will always be some completely random stuck protons or hard-to-evaporate water.

But I suspect this can be done, because here is a system with an exactly known macroscopic number of atoms. Perhaps if you centrifuge the DNA and compare the density of water to the density of DNA-water?

Question: Can you define the kg with PCR?

share|cite|improve this question

Short: Rereading the question, keeping the atomic weight of all these different (types of) atoms the same is kind of impossible.

I think not, or at least not practically. The problem is that many of the molecules in DNA have several isotopes which occur in nature. Therefore the mass of the dna bunch would differ from production to production. Defining it as the average of the production or by specifying the normal isotope ratios would make it a 'sloppy' definition. There are other atoms where the different isotopes are easier (instead of almost impossible) to separate and therefore are more useful to make an exact kg.


Short: PCR is not perfect.

Furthermore, how do you know the PCR did in every step multiply the number of sequences by exactly two and didn't skip one. Or did it incidentally introduce an extra erroneous loop or skip the begin.

Conclusion: While it might be possible to do so, it would be higly impractical, as you would never make an exact one kilogram as easily as you could make one from say silicon and probably carbon-12 is even easier.

share|cite|improve this answer
I think this is not right--- the isotope issue is for sure not important, you just use the natural isotope average, the statistical fluctuations are usually negligible at the level of accuracy of a mass standard (the same issue also negligibly affects the standerd kilograms), as for the imperfections in PCR, I believe they can be controlled, I think you can get reliable exponential amplification of a relatively big strand with the right density of reagents. – Ron Maimon Jul 17 '12 at 14:45

According to the error rate in the PCR reaction is about 1 error for every 9000 nucleotides replicated. I suspect this makes PCR too unreliable to use as a standard even if you have isotopically pure feedstuffs.

share|cite|improve this answer
YOu don't need isotopic purity, and if you have a C-G strand, some errors don't matter. If you have an inserted base, that matters, but I don't think that happens even 1 in a billion billion, it makes the two strands mismatch in number. THe error is for complementary codon copying, and this fixes itself in a pure CG enviroment. – Ron Maimon Jul 17 '12 at 16:47
I checked the Taq polemerase, the errors of 1:9000 are of the type that don't matter for C-G random replication (no AT), the error would be swapping a G for a C, and this error doesn't amplify if the bases are made correct at the end. The issue is whether you can assure there are no G-G pairs of C-C pairs at the end. – Ron Maimon Jul 18 '12 at 0:52

The "PCR is not perfect" issue has an additional complication in that errors, particularly early on in the cycle, are exponentially amplified.

You never know when the PCR enzymes might make a mistake, or some cosmic ray will hit your molecule and render it uncopiable, or whatever. Further, if you plan on making macroscopically many copies, then you're going to have a huge number of copies to keep track of and make sure are OK. Then since the total mass is an exponential times the initial mass, errors in the initial mass are exponentially amplified. Thus a one-molecule cosmic-ray hit at the 100-molecule stage will translate into a 1% error, which is way below the ten significant figures you'd need to match IPK-based standards.

share|cite|improve this answer
You have to sheild from radiation for sure. I suggested using C-G pairs, because if you only have these, any error of replacing is automatically mass-neutral. – Ron Maimon Jul 17 '12 at 16:46
It's not the specific sources of noise that bug me but the scaling on top of that noise. – Emilio Pisanty Jul 17 '12 at 23:46

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.