$25 PCR Thermal Cycler
Norman Wang
Abstract:
This article describes the construction of a safe, programmable,
automatic thermal cycler for PCR that can be easily constructed by
persons with basic soldering and mechanical skills for under $25 in
parts and a modest computer such as IBM 486, all of which are readily
available. The cycler relies on the heating provided by an
incandescent light bulb and cooling by simple convection.
Blais, Brian S. and Betsch, David F., "A Programmable $25 Thermal
Cycler for PCR" (2002). Science and Technology Faculty Journal
Articles. Paper 6.
via http://digitalcommons.bryant.edu/sci_jou/6
Tito Jankowski
Bryan Bishop
> Is anyone building a thermal cycler like this?
$25 is a bit too much. How about $10? :-)
Gene Hacker
On Feb 19, 1:28 pm, Norman Wang <analo...@gmail.com> wrote:
> The cycler relies on the heating provided by an
> incandescent light bulb and cooling by simple convection.
call it an Easy-PCR Thermocycler.
Regarding, the temperature resistant container, how about using a
plant pot instead of a stainless steel pot?
How about a $1.99 Ikea BETA plant pot? http://www.ikea.com/us/en/catalog/products/50072320
It has a smaller diameter, so maybe the $4.99 BETA plant pot with a
handle? http://www.ikea.com/us/en/catalog/products/00069396
Now it only costs $19 - $22.
Tito Jankowski
Bryan Bishop
> A PCR machine costs $3000+ new
What for? We're suggesting versions that would cost less than 1%-
hell, somebody throw together a grant application, get on it. :-)
> A gel box costs $250-$500
How is it worth $250 ? Really? Is that including the power supply?
> An illuminator for seeing DNA in a gel box costs
> $1135: http://tinyurl.com/dzfhwd
So, what about a raster scanning UV LED and diode setup?
Cory Tobin
>> A PCR machine costs $3000+ new
>
> What for? We're suggesting versions that would cost less than 1%-
> hell, somebody throw together a grant application, get on it. :-)
>
>> A gel box costs $250-$500
>
> How is it worth $250 ? Really? Is that including the power supply?
No power supply. I recently got a new one from BioRad. Medium sized.
300 bucks. It is pretty excessive for a few pieces of acrylic and
some platinum wire. But as long as academic and commercial labs are
willing to pay the price rather than build their own boxes, BioRad is
going to charge as much as they can.
It really comes down to a matter of time. Most university biology
labs would rather spend the money and save the time rather than spend
the time and save the money.
-Cory
Gene Hacker
On Feb 19, 6:24 pm, Tito Jankowski <titojankow...@gmail.com> wrote:
> For those who need some perspective:
> A PCR machine costs $3000+ new
>
> A gel box costs $250-$500
>
> An illuminator for seeing DNA in a gel box costs $1135:http://tinyurl.com/dzfhwd
>
light or a UV disinfector.
UV disinfector 185-365 nm, $139.99, but I think one might be able to
find a cheaper one:
http://www.activeforever.com/p-2128-zadro-nano-uv-wand-disinfection-light-scanner.aspx
Paul Anderson
>
>
>
> I'm willing to bet we could make an illuminator using a tanning bed
> light or a UV disinfector.
>
--
Paul Anderson
VE3HOP
wacky...@gmail.com
http://www.andersonloco.com
QRP ARCI #13228, GQRP #12447
Josh Perfetto
Much safer and easier to work with, and no hazardous waste.
-Josh
Gene Hacker
>What wavelength do you need? There are UV LEDs available now.
LEDs might be worth considering
http://www.nichia.com/product/led-lamp-uv.html
NOTE: this is a repost, because I didn't know that reply to author
meant that only the author could see the post.
Gene Hacker
On Feb 19, 7:45 pm, Josh Perfetto <j...@snowrise.com> wrote:
> If you use SYBR Safe for staining you can use a visible 502nm wavelength.
> Much safer and easier to work with, and no hazardous waste.
>
> -Josh
>
much does it cost? 502 nm is easy, heck, I even have a laser close to
that.
Cory Tobin
> much does it cost? 502 nm is easy, heck, I even have a laser close to
> that.
For SYBR dyes I usually stain the gel afterwards rather than mixing
the dye into the gel. If the solution it was sent in says "1000x"
then first you need to dilute it down to 1x. So if you think 10mL is
enough to submerge your gel, then mix 10uL of the dye into 10mL of
water. Then put the gel in some sort of container and pour the 1x dye
on the gel (completely submerging it). Let it sit in the dark for ~15
minutes. Take the gel out and transilluminate.
-Cory
Thomas Knight
buffer, not water. The gel lanes will distort due to buffer
equilibration if you put a gel into water. The stain solution can be
reused several times.
Josh Perfetto
products as well (SYBR Green etc). You pretty much use it just as you would
EtBr. The easiest way is to add it to the agarose solution before the gel
solidifies. To see the bands after running the gel, put it on the blue
502nm or so (502nm is the excitation peak, but other wavelengths are usable
like 470 nm LEDs) transilluminator. SYBR Safe is florescent and will emit
in the orange spectrum, so you will need to place an orange filter over the
gel in order to see it. Alternatively if you want to photograph it, you can
use an orange filter for your camera, or if you want to cut bands out, you
can wear orange glasses. Most of the blue light from the illumination
source will be filtered out by the orange filters, making it very easy to
see the bands.
With EtBr and a UV light source, you will pretty much go through the same
procedure, except that now you have to be careful about the UV light harming
your eyes/burning your skin, which makes it more of a pain to work with.
EtBr is also a hazardous waste in many municipalities, and so you can have
disposal issues.
Being sold by Invitrogen though (
http://www.invitrogen.com/site/us/en/home/Products-and-Services/Applications
/Nucleic-Acid-Purification-and-Analysis/Nucleic-Acid-Gel-Electrophoresis/nuc
leic_acid_gel_electrophoresis/Nucleic-Acid-Stains/SybrSafe-A-better-DNA-gel-
stain.html) you'll need a company address for shipping. However you'd
probably have the same troubles obtaining EtBr.
-Josh
Josh Perfetto
http://www.labsupplymall.com/nucleic-acid-stains-84/gr-safe-nucleic-acid-sta
in-500-ul-290.html
BTW I know a lot of people say that with SYBR Safe you should do
post-staining instead of adding it to the gel. I've always added it to the
gel and never had any problems. Not really sure what the issue is.
-Josh
Cory Tobin
> http://www.labsupplymall.com/nucleic-acid-stains-84/gr-safe-nucleic-acid-sta
> in-500-ul-290.html
>
> BTW I know a lot of people say that with SYBR Safe you should do
> post-staining instead of adding it to the gel. I've always added it to the
> gel and never had any problems. Not really sure what the issue is.
SYBR Gold and SYBR Green seem to make the bands run slower. I've
never used SYBR Safe though. I just assumed it was the same.
-Cory
Knut Norstog
For this design, wouldn't the flouropolymer tubing need to be changed
per use? Also, my feeble mind can't seem to grasp how this could work
for 100 [give or take] samples at one time.
I just started at a shrimp pathology lab that uses PCR based
diagnostics (we get a lot of sample at times).
Thanks
On Feb 19, 1:10 pm, Bryan Bishop <kanz...@gmail.com> wrote:
> > Is anyone building a thermal cycler like this?
>
> $25 is a bit too much. How about $10? :-)
>
> http://groups.google.com/group/diybio/msg/3cc48f8503d133de
>
> 1 512 203 0507
Norman Wang
yet but maybe it will soon. Current LED UV bulbs have wavelengths
centered around 400. It could possibly work, not sure how wide is the
wavelength distribution.
On Feb 19, 3:38 pm, Paul Anderson <wackyvor...@gmail.com> wrote:
> On Thu, Feb 19, 2009 at 8:36 PM, Gene Hacker <cam...@gmail.com> wrote:
>
> > I'm willing to bet we could make an illuminator using a tanning bed
> > light or a UV disinfector.
>
> What wavelength do you need? There are UV LEDs available now.
>
> --
> Paul Anderson
> VE3HOP
Thomas Knight
that matter, desirable). The new gel box design will function
perfectly with EtBr using blue LEDs.
peter
can take picture with regular camera (cell phone camera if you like)
P
Bryan Bishop
> SYBR green/gold too can be visualized with blue light, and also one
> can take picture with regular camera (cell phone camera if you like)
So if that's true, if anyone likes doing cell phone application
development (meh, I stay away from it, just because I have better
things to do), it would be neat to be able to take a photo and then
analyze the migration distance with respect to some marker in the
photo, which is a simple image processing or analysis trick. At the
very least, this could be something submitted to the bioperl project,
or bioruby project or something, using programs like imagemagick.
bioperl
http://bioperl.org/
http://www.bioperl.org/wiki/History_of_BioPerl
imagemagick
http://imagemagick.org/
"ImageMagick(R) is a software suite to create, edit, and compose
bitmap images. It can read, convert and write images in a variety of
formats (over 100) including DPX, EXR, GIF, JPEG, JPEG-2000, PDF,
PhotoCD, PNG, Postscript, SVG, and TIFF. Use ImageMagick to translate,
flip, mirror, rotate, scale, shear and transform images, adjust image
colors, apply various special effects, or draw text, lines, polygons,
ellipses and Bézier curves."
Oops. Turns out somebody beat us to it.
http://open2dprot.sourceforge.net/Flicker/
"Flicker is an open-source stand-alone computer program for visually
comparing 2D gel images. Two-dimensional polyacrylamide gel
electrophoresis (2D-PAGE) gels are often difficult to compare because
of rubber-sheet distortions. Flicker allows you to visually compare
your gel images against each other or against those found in Internet
databases. Many published Internet gels have a subset of spots
identified which may make them useful to compare with your gels. Some
of these Internet gels are active maps that you can click on a spot to
inquire of its identity. You may be able to draw putative conclusions
as to the identification of some spots in your gels that visually
appear to be the same spots as in reference gels. The Flicker program
integrates these various needs to help you try to make putative spot
identifications. See the Quick start examples below for a short list
of some of these methods."
Now just throw that on to a cell phone app, and suddenly thousands of
biology students all over the worl-- ah, pipe dreams.
- Bryan
Guido D. Núñez-Mujica
for the iPhone, but for Java phones, thousands around the world would
be happy. It would simplify the analysis of samples.
On the other hand, I am still interested in using cell phones as Elisa
plate readers, but I am not sure of how feasible it is.
Meredith L. Patterson
> So if that's true, if anyone likes doing cell phone application
> development (meh, I stay away from it, just because I have better
> things to do), it would be neat to be able to take a photo and then
> analyze the migration distance with respect to some marker in the
> photo, which is a simple image processing or analysis trick. At the
> very least, this could be something submitted to the bioperl project,
> or bioruby project or something, using programs like imagemagick.
Nokia S60-series phone (comes with a Python interpreter) + biopython +
Python Imaging Library = win.
Sigh. I've been meaning to get one of those phones for a while now, I
suppose this is just one more reason.
--mlp
JonathanCline
wrote:
like:
- grey scaling
- sharpening
- edge detection
- rotation
- threshold
The above image processing algorithms are very simple and would amount
to under 500 lines of well-spaced C code. Note: No JPEG libraries
required since it should use the raw bitmap. In comparison, python
interpreter+related libraries+kitchen sink GNU library+imagemagik
addon library+etc,etc are >100k lines of code. Not to mention
slow. As an iPhone app, it would be very simple using all standard
libraries. Is a 2 Megapixel camera sufficient to differentiate/
discriminate the lines?
Originally I was thinking about using a phototransistor approach for
building a read-head which could be manually scanned or motor-scanned
across the gel. It's pretty much the same concept as the line-
following robots I used to build, except in reverse. The line-
detection side is very easy - even adding a bluetooth transmitter to
it for wireless operation, it's still easy (serial output to bluetooth
dongle). There is no need to take a digital image of the gel from the
data processing standpoint, except for the warm fuzzies or glam cover
publishing in Nature.
## Jonathan Cline
## jcl...@ieee.org
## Mobile: +1-805-617-0223
########################
Dan
doesn't work very well as a prestain in gels. For one, it's fairly
pricey compared to ETBr so staining entire gels gets expensive
quickly. Preloading the DNA with SYBR Green doesn't work either since
different bands run at different speeds depending on the SYBR Green/
DNA ration in each sample. I found that loading a bunch of junk DNA
in each well (salmon spern DNA at least 10X the amount of your sample)
tends to renomalize everything but ends up being too much effort to be
worth it.
Now, that's assuming you're pouring full gels (usually 30-50mL per
gel). If you pour mini slab gels (I mentioned this in the precast
gels thread), the total volume per gel is low enough that the SYBR
Green costs get fairly reasonable for prestaining the entire gel. I
haven't tried entire gel prestaining with SYBR Green myself but as
long as all the DNA is slowed equally, it shouldn't cause a problem.
ETBr also retards DNA movement though a gel so this is not an
insurmountable issue. If anyone is set up to test this, running DNA
ladder at different concentrations and checking to ensure that all the
bands run at the same rate should be a decent check for the
practicality of this.
Bryan Bishop
> On Feb 21, 8:26 am, "Meredith L. Patterson" wrote:
>> On Sat, Feb 21, 2009 at 4:03 AM, Bryan Bishop wrote:
>> > So if that's true, if anyone likes doing cell phone application
>> > development (meh, I stay away from it, just because I have better
>> > things to do), it would be neat to be able to take a photo and then
>> > analyze the migration distance with respect to some marker in the
>> > photo, which is a simple image processing or analysis trick. At the
>> > very least, this could be something submitted to the bioperl project,
>> > or bioruby project or something, using programs like imagemagick.
>>
>> Nokia S60-series phone (comes with a Python interpreter) + biopython +
>> Python Imaging Library = win.
>>
>> Sigh. I've been meaning to get one of those phones for a while now, I
>> suppose this is just one more reason.
>
> I'm not sure the algorithms used however I imagine it's something
> like:
>
> - grey scaling
>
> - sharpening
>
> - edge detection
>
> - rotation
>
> - threshold
Surely there's a paper on this out there?
> The above image processing algorithms are very simple and would amount
> to under 500 lines of well-spaced C code. Note: No JPEG libraries
> required since it should use the raw bitmap. In comparison, python
> interpreter+related libraries+kitchen sink GNU library+imagemagik
> addon library+etc,etc are >100k lines of code. Not to mention
> slow. As an iPhone app, it would be very simple using all standard
> libraries. Is a 2 Megapixel camera sufficient to differentiate/
> discriminate the lines?
Can somebody check please? The next gel you do, snap a few photos, or
take a few photos out from your archive.
> Originally I was thinking about using a phototransistor approach for
> building a read-head which could be manually scanned or motor-scanned
> across the gel. It's pretty much the same concept as the line-
> following robots I used to build, except in reverse. The line-
> detection side is very easy - even adding a bluetooth transmitter to
> it for wireless operation, it's still easy (serial output to bluetooth
> dongle). There is no need to take a digital image of the gel from the
> data processing standpoint, except for the warm fuzzies or glam cover
> publishing in Nature.
I'd be concerned about manual scanning- there's no way you're going to
be able to do that at a reasonable constant rate with your hands. And
the motor-scanning mechanism is also kind of complex to set up in this
case. The cell phone idea, I think, is a cheap solution. But ideally
one that doesn't require a cell phone is even better. The cheap
image-bed scanners from garage sales, perhaps. Any way you can get the
data into a camera, or scanner, I think, will be sufficient.
JonathanCline
> On Mon, Feb 23, 2009 at 2:49 PM, JonathanCline wrote:
>
> > I'm not sure the algorithms used however I imagine it's something
> > like:
>
> > - grey scaling
>
> > - sharpening
>
> > - edge detection
>
> > - rotation
>
> > - threshold
>
> Surely there's a paper on this out there?
processing required is not complex and image processing runs in tight
loops (for i=0..799, for j=0..599, calculate something, assign
something, endfor, endfor). The algorithms are in standard grad level
image processing books, or maybe one of the "algorithms in C" type
books. Mostly, the algorithms are tweaked experimentally. Maybe with
a little MATLAB simulation prior to the C.
For articles - see below. Unfortunately one of the articles mentions
convolution, which is slow.. but perhaps this could be worked around
with analog processing.
> > Originally I was thinking about using a phototransistor approach for
> > building a read-head which could be manually scanned or motor-scanned
> > across the gel. It's pretty much the same concept as the line-
> > following robots I used to build, except in reverse. The line-
> > detection side is very easy - even adding a bluetooth transmitter to
> > it for wireless operation, it's still easy (serial output to bluetooth
> > dongle). There is no need to take a digital image of the gel from the
> > data processing standpoint, except for the warm fuzzies or glam cover
> > publishing in Nature.
>
> I'd be concerned about manual scanning- there's no way you're going to
> be able to do that at a reasonable constant rate with your hands. And
> the motor-scanning mechanism is also kind of complex to set up in this
> case. The cell phone idea, I think, is a cheap solution. But ideally
> one that doesn't require a cell phone is even better. The cheap
> image-bed scanners from garage sales, perhaps. Any way you can get the
> data into a camera, or scanner, I think, will be sufficient.
pinwheel for measuring distance. It could also include some simple
mechanical rails for alignment. So whatever speed you move it, even
forward/backward, it always knows where it is. It's quite simple
actually; I can imagine a solution using an audio CD as a wheel with
calibrated markings (usually drill holes), this could have mm
resolution. For a motor solution, a belt drive could ensure reliable
positioning up/down the lane, and the user could position the head by
sliding it along rails across the lanes. Very simple!
Articles that I just found:
Software for Automated Analysis of DNA Fingerprinting Gel
http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=12727910
"Here we describe software tools for the automated detection of DNA
restriction fragments resolved on agarose fingerprinting gels. [...]
Sulston et al. (1988, 1989) were among the first to develop methods
for automating bandcalling. They developed software for lane tracking
and band detection that was the precursor to the IMAGE package, which
is available from the Sanger Institute (http://www.sanger.ac/Software/
Image). [...] Gel (TIFF) images are collected on a Molecular
Dynamics Fluorimager. Although the Fluorimager is capable of different
settings, BandLeader has been tuned to use 200-micron-square pixels.
Each gel image is 1000 × 1200 pixels. The gel image is partitioned
into 121 single-lane images, each 1000 pixels long × 9 pixels wide, as
a result of the lane tracking process in IMAGE, but is otherwise
unprocessed prior to our analysis."
A Software System for Data Analysis in Automated DNA Sequencing
http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=9647639
"Two primary software packages were developed as part of this effort.
One package, called GelImager, is aimed at gel image analysis. It
allows the viewing and manipulation of the electrophoretic image files
collected from the electrophoresis instrument, acts as an interface to
the lane-finding software Getlanes which was developed by the St.
Louis Genome Center (Cooper et al. 1996), and allows the retracking of
lanes determined by that software (those who wish to use Getlanes need
to obtain permission from the authors of that package). A second
software package, called BaseFinder, is designed to process the trace
files output by the lane-finding module, performing analysis and base-
calling to produce a DNA sequence from them. [...] All of these
programs and source code are available for free for noncommercial use.
More information can be obtained from our web site (http://
smithlab.chem.wisc.edu/Research/informatics.html). They were developed
using the OpenStep object toolkit (http://www.enterprise.apple.com/
openstep). Typical processing of a trace on a Pentium-Pro 200-Mh Z
class computer takes 1 min per lane. A large portion of this time is
spent in the deconvolution process. Although time consuming, the
latter appears to account for an important portion of the gain in
accuracy achieved by BaseFinder’s preprocessing. It also appears that
the deconvolution routines can be further optimized, and computers
continue to get faster, therefore this is not perceived as a major
stumbling block to the program’s use in production sequencing. During
these efforts, it was observed that typical useable read lengths are
in the range of 700–750 bases."