According to Essential Science Indicators
fromThomson
Reuters, the paper "Analysis of relative
gene expression data using real-time quantitative PCR
and the 2−ΔΔCTmethod,"
(Livak KJ and Schmittgen TD, Methods 25[4]:
402-8, December 2001) is ranked at #3 among Biology
& Biochemistry papers published in the past decade,
with 4,137 cites.
The paper's authors, Dr. Kenneth Livak and Dr.
Thomas Schmittgen, recently spoke with ScienceWatch.com
correspondent Gary Taubes about this paper and its
impact on academia and industry.
Dr. Livak is currently in Research and Development at the Fluidigm
Corporation in South San Francisco. Dr. Schmittgen is an Associate
Professor at the Ohio State University College of Pharmacy.
Below, the paper's authors,
Dr. Kenneth Livak and Dr. Thomas Schmittgen, talk with
ScienceWatch.com correspondent Gary Taubes about this paper and
its impact on academia and industry.
Ken, what motivated the work that led to the
2−ΔΔCT method for real-time PCR and your 2001
Methods paper?
Livak: In 1994, I went to work for Applied Biosystems, which had just
started this project on real-time PCR. I was part of the team there that
developed the first instrument for doing it. These projects were
multidisciplinary in that they had engineers, chemists, biologists, etc.,
and I took the lead position on the molecular biology part in terms of
developing assays, optimizing assays, and getting detection with the probes
to work well. Because Applied Biosystems was introducing a whole new
system, they needed documentation to teach their customers how to deal with
the results. We provided a document called User Bulletin Number 2. I was
the main person involved in drafting that User Bulletin, collecting all the
data, and getting it published.
"The
number of publications that cite
real-time PCR has probably been
growing exponentially over the past
decade."
~Thomas Schmittgen
This bulletin addressed the problem of quantitation by real-time PCR, which
is different conceptually than the way people usually think about
quantitation. Naively, they think you do a fluorescence measurement and the
amount of fluorescence will tell you how much analyte is there. But what
you get out of real-time PCR is really more of a kinetic measurement. You
do your quantitation based on how many cycles it takes to reach a certain
fluorescence threshold. So the whole way you do the quantitation was
different than the way people were thinking about it in the past. It really
required some basic mathematics background so people could readily take the
real-time PCR data and get quantitative results out of it.
If this started as a User Bulletin for an Applied
Biosystems product, how did you two come to collaborate on the
Methods paper?
Schmittgen: Well, Ken developed this method and published it in the User
Bulletin. So when you purchased the instrument from Applied Biosystems,
that User Bulletin would come with all the various manuals. This was
pre-internet days, so it was mostly only available to people who had the
instrument. Anyway, that was published, but it was not in the general
literature. Early on, when people needed to cite something in a paper, they
would have to cite the User Bulletin from the company. I was working on a
paper for Analytical Biochemistry and I, too, cited the Applied
Biosystems User Bulletin, and one of the comments that came back from the
reviewers was that it was not appropriate to cite a User Bulletin in the
literature.
It just so happens that at the time I was editing a thematic issue for the
journal Methods on quantitative real-time PCR. I said, "Why
don’t I publish Livak’s derivation of the equation in this
journal? That way it would be available to everyone in the literature." I
had never met Ken, but I called him up, told him what I wanted to do and he
agreed with it, and that’s what we did.
Livak: You have to realize that when you’re working
at a company like Applied Biosystems, there isn’t quite the same
incentive to publish in academic journals that there is in the academic
world. It’s a good thing to publish, but it’s kind of
secondary. It’s not publish or perish in my world. We were interested
mainly in designing the instruments and the reagents and developing the
methods. Until Tom called, I just wasn’t motivated to submit this to
a scientific publication. With Tom offering to help, it got me over the
activation barrier of doing the work necessary to make this an academic
publication.
Naïve question perhaps, but what’s the
difference between real-time PCR and not real-time, or regular
PCR?
Schmittgen: Regular PCR amplifies the RNA or DNA in a sample, and real-time
PCR amplifies it and quantifies it. You’re actually measuring
something as it occurs in real time. Following the PCR, you get a number,
and that number tells you how much RNA or DNA is in the reaction.
How hard was it to come up with the mathematics and make
this work?
Livak: Developing the real-time PCR probe and the system itself was the
hard part, and that’s what enabled this technology to become
widespread. The mathematics does take careful thought, but our primary
advantage was that we were the first ones coming out with the system and
the first ones doing the math. A lot of people could have done what we did;
we just got there before them.
What issues of real-time PCR does this paper address or
solve?
Schmittgen: What it does, in its simplest form, is allow you to essentially
work with the data that is generated from the experiment. You’re
getting numbers from the computer that’s attached to the instrument.
The question is, what do you do with these numbers? How do you use them?
How do you present the data? What Ken’s equations do is allow you to
present the data in a way that makes sense of it.
Basically what you’re doing is looking at the gene expression, the
RNA; you take a cell, treat it with a drug; take another cell, use it as
the control, and then you measure the amount of RNA in both cells. This
equation then allows you to relate the expression of genes in the treated
cell to the untreated cell. What Ken had done very elegantly was to come up
with the equations and various assumptions that allowed him to derive the
equation. It was a beautiful piece of mathematics. Part of the reason why
the method is so popular is that it’s very simple and yet based on
this theory of PCR amplification.
Livak: The reason that the method—this
2−ΔΔCT method—works is actually pretty
interesting. Whenever people do the mathematics for these sorts of things,
you have to make all these assumptions so that you can get a relatively
straightforward mathematical expression. It turns out that the actual
process of PCR is much more complex than these assumptions would indicate,
but the simple approximation that we made in the paper still manages to
give you very good data. That’s fascinating to me. How is it that
these mathematical expressions can sometimes be very useful, even though
we’ve come to understand the underlying processes are far more
complicated? It turns out that our original assumptions may not be
precisely accurate, but there’s still something about the overall
process that enables the equation to provide good quantitative results.
How much did the Methods article actually
differ from the User Bulletin?
"It turns out that the
actual process of PCR is much more complex than
these assumptions would indicate, but the simple
approximation that we made in the paper still
manages to give you very good
data."
~Kenneth Livak
Schmittgen: Probably 70% of it was more or less the same. We didn’t
just cut and paste the bulletin into the paper. The derivations of the
equations were identical, but we presented it in a more educational manner
so people could learn how to use it. We provided different examples and a
couple of different modifications of the method.
How did the writing process go, considering you guys had
never met before and didn’t even know each other until
Tom’s phone call?
Livak: We basically started with the User Bulletin, and then Tom put that
into the form of an academic paper, the way he saw it. Then I worked on
writing some of it and having it flow as a paper. Tom added material,
including some additional experiments. So he did the first draft based on
my User Bulletin, and we then collaborated from there to get it to final
form.
Are you surprised at how influential your method has
been and how remarkably highly cited the Methods paper has
become?
Schmittgen: I’m very surprised. As I said, my only intention was that
I was working on this journal issue and I thought it would make sense to
include this paper in the issue and give investigators like myself
something in the archival literature to cite. I had no idea it would go so
far.
Livak: I’ve only recently appreciated the place this
real-time PCR work has taken me. When you’re in the trenches working
on it, getting the system up and out there, you have one view. And once you
introduce these systems, it takes a certain amount of time before they
start getting used widely and really start having influence. It’s
only over the past two or three years that I’ve become aware how
pervasive real-time PCR has become. In that context, I’m not that
surprised the paper has been cited a lot. What’s in the paper is
pretty basic and fundamental to the technology. It’s really the
foundation of how you do things in real-time PCR; it’s the sort of
thing that people would refer back to, the place you start if you want to
get meaningful quantitative results.
How much has real-time PCR changed since 2001? Is it
much more sophisticated?
Livak: Actually, not in any fundamental sense. What’s changed is that
people are doing things faster, and with higher throughput. The actual
underlying technology—how you set up the reactions and do the
fluorescence measurements—is still fundamentally what it always was.
Schmittgen: One thing that’s changed is that
it’s become far more popular to use the technology. The number of
publications that cite real-time PCR has probably been growing
exponentially over the past decade. And more people are using the
technology to study more genes. This is the high-throughput aspect of it.
Instead of just looking at one gene now, people are looking at
hundreds.
Kenneth J. Livak, Ph.D.
Fluidigm Corporation
South San Francisco, CA, USA
Thomas D. Schmittgen, Ph.D.
College of Pharmacy
Ohio State University
Columbus, OH, USA
Dr. Kenneth Livak &
Dr. Thomas Schmittgen's
most-cited paper with 4,137 cites to
date:
Livak KJ and Schmittgen TD, "Analysis of relative gene
expression data using real-time quantitative PCR and the
2−ΔΔCT method,"Methods
25(4): 402-8, December 2001. 4,137 cites. Source:
Essential Science Indicators from
Thomson
Reuters.