qPCR fluorescence detection dynamic range

[Josh Perfetto]

Hello,

Does anyone have a sense of what real time PCR instruments typically offer for the dynamic range and resolution of the fluorescence detection (and NOT the dynamic range of the overall detection)? I am trying to determine the minimal requirements for a cheaper machine.

-Josh

[Josiah Zayner]

http://lmgtfy.com/?q=fluorometer

Click first link

[Josh Perfetto]

Josiah,

I don't think you have understood what I was asking. My question is not about the dynamic range of purpose-built fluorometer units, but of the dynamic range of the fluorescence detection that is incorporated into real time PCR systems. I suspect the range is far less than the qubit, for example.

-Josh

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.
To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/EIud2NaEehQJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Nathan McCorkle]

Why would you suspect that? It /is/ simply a fluorometer built in to a
PCR machine... usually a CCD or CMOS array sensor is used to get all
the wells synchronously and without optical switching

--
-Nathan

[Nathan McCorkle]

Can you describe fluoro and overall detection meanings? I'm not seeing
where they would be separated in that instrument... total illumination
wouldn't be much different, you need that filter in-line to remove the
exciter radiation (the light which excites the fluorophore, I may have
just made that term up)


The dynamic range will simply be the min and max mass of DNA that's
detectable, given some fixed volume of reaction. That's gonna be
determined by the intended use of the instrument. Generally I think
people want to know if there's at least one molecule of DNA being
produced, so you need to take your base fluorescense and subtract it
from all subsequent readings... so now you need to experimentally
determine efficiency of (or correlate)
DNA-molecules-in-machine-with-dye and ADC readings out of imager. For
really good dyes this might come with the spec sheet, but it's likely
buried beneath licensing and you-don't-need-to-know... for the cheaper
dyes (gelgreen, gelred) they simply use somewhat nonsensical terms
without proper units ("10ng detection limit") which aren't useful
because the square area (FOV, field of view) matters and is linked to
the quantum efficiency of the imager.

If you simply use the gelgreen spec (which doesn't include all needed
units) you can make some interesting extrapolations, assuming
'standard' gel well dimensions.

[Josiah Zayner]

I think what I was trying to get at was...

What type of measurement are you looking for? In photon counts? Arb. units? Intensity?
Do you want the arb. unit range? The concentration of DNA?
What fluorophores are you using?
Integrating over what time range?

What is the fluorescence intensity of a qubit?

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

[Josh Perfetto]

I suspect that because it would be too expensive and excessive. The Qubit claims a 5 order of magnitude dynamic range. Lets take that to mean 100,000. That would require at least a 17 bit CCD. That would be rather costly and I can't see why such resolution would be needed, because most of your dynamic range in the overall qPCR detection is coming from the cycle count.

On the other hand, I am not sure what the practical lower range is. That's why I'm asking what's typical.

-Josh

[Josh Perfetto]

qPCR produces a dynamic range on the order of 7-10 logs. No analog sensor I know of comes close to that. The way qPCR works is you don't quantify your DNA based only on the fluorescence measurement, but also based on the cycle count when that measurement was taken. Since your signal is doubling each cycle assuming 100% efficiency, you effectively get this huge dynamic range, even though your fluorometer comes no where near that range.

Josiah, dynamic range refers to the ratio of the largest and smallest measurable quantity. Thus you could express this as an integer, log number, bit count, etc. qPCR manufacturers usually publicize the dynamic range of what you can detect (i.e. you could detect template in a sample of concentration X, and template in a sample of concentration X * 10^8), but not the dynamic range of the fluorometer (at least I haven't found it after hours of googling). It's easy to find the dynamic range of a fluorometer like the qubit since it's a major selling point, but in this case it seems more of a detail of the machine.

-Josh

[Josiah Zayner]

Can't one theoretically just take any analog sensor and connect it to a 24bit ADC and you immediately have 7 logs?
Also, PMTs have huge dynamic ranges don't they? And they use them in Real-Time PCR.

I thought this was more hardware/software thing then a sensor thing? Noise, &c.

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

[Josh Perfetto]

You are correct that for reasons like noise (which would vastly exceed signal at the PCR starting concentrations) this doesn't work out. Thus you have to look at the complete system and not just one component like an ADC. The basic idea behind real time PCR is that you couple your detector with an exponential amplification process and measure the signal throughout that process. That's why it's called "real time".

-Josh

[Nathan McCorkle]

You're not taking into account exposure time, I believe. 24-bit adcs are common but you need impecable circuit design to achieve good readings.

[Josh Perfetto]

I finally found something - the BioRad MyiQ and iQ5 have only a 12-bit (4096 dynamic range) CCD camera: http://www.bio-rad.com/webroot/web/pdf/lsr/literature/Bulletin_5462.pdf (page 4).

-Josh

[Nathan McCorkle]

Also some ADCs have changable amplifiers builtin, so combined with an electronic shutter you still have more dynamic range

[Jelmer Cnossen]

The same PDF also says "Sensitive charge-coupled device (CCD)-based optics enable accurate quantitation over a dynamic range of 9 orders of magnitude". 

[Patrik D'haeseleer]

On Tuesday, May 14, 2013 12:37:39 AM UTC-7, Jelmer Cnossen wrote:

The same PDF also says "Sensitive charge-coupled device (CCD)-based optics enable accurate quantitation over a dynamic range of 9 orders of magnitude". 

Yeah, that's what Josh meant when he mentioned "sensitivity of the overall detection". Since you're doubling concentration at every round of the PCR, and 10^9 is about equal to 2^30, this qPCR can measure initial DNA concentrations that differ by about 30 PCR cycles in when they reach the detection threshold.

However, Josh's question was about the typical resolution with which the intensity is measure at each cycle. In principle, you could build a qPCR that uses a 1-bit threshold detector yet still has an overall dynamic range of 9 orders of magnitude: just keep doing PCR cycles until your detector flips from "below threshold" to "above threshold". But that would only allow you to measure the initial concentration to within the nearest power of 2. Apparently, the BioRad MyiQ and iQ5 use a 12-bit sensors, but their overall dynamic range is far more than 12 bits.

There's an interesting similarity here with how floating point numbers are stored in binary represeantation on a computer. Essentially, Josh was asking how many bits are typically allocated to storing the mantissa of the qPCR floating point, rather than the exponent. :-)

[Josh Perfetto]

On Tue, May 14, 2013 at 2:59 PM, Patrik D'haeseleer <pat...@gmail.com> wrote:
> There's an interesting similarity here with how floating point numbers are
> stored in binary represeantation on a computer. Essentially, Josh was asking
> how many bits are typically allocated to storing the mantissa of the qPCR
> floating point, rather than the exponent. :-)

Wow that's a really good way to think about it. This suggests the
problem could be worked backwards from specified resolution values.

I noticed the BioRad CFX96 literature shows amplification curves with
text like "Average Ct 19.81 +/- 0.10". To me that implies 0.01
resolution in the Ct value, implying at least 7 bits in the mantissa,
though I'm not sure if there are practical issues requiring more.

If an 8 bit CCD could be used, it is very interesting, as these are
much cheaper.

-Josh

[Nathan McCorkle]

CCDs are analog, not digital. Check out the ADC I'm using for
openSpectrometer, it's 16-bit with a pre-amp and all the correlated
double sampling circuitry built-in, around 6 MSPS at ~$4

http://openspectrometer.com/datasheets/WM8253.pdf

[Josh Perfetto]

What is the difference between an 8-bit and 16-bit camera then? Are
you saying you could just switch the ADC and use the same CCD?

-Josh

Sent from my iPhone

[Nathan McCorkle]

Yes, CMOS imaging chips often have all the digitization circuitry
built-in, which is why they've become so cheap. This reduces the
quantum efficiency , but leads to streamlined fabrication and PCB
design.

[Josh Perfetto]

But don't you have an issue with noise limiting your dynamic range if
you just take a camera designed for 8 bits and put a 16 bit ADC on it?
Like say your CCD holds 0-100,000 electrons. That's more than your 16
bit ADC so the ADC is your limiting factor. But then you get 50
electrons of noise in each pixel, so now your dynamic range drops to
2000, so you are effectively getting just an 11 bit camera? (Or even
worse, just some example numbers)

-Josh

[Nathan McCorkle]

Noise needs to be dealt with properly, see at least pg 4 of this PDF:
http://www.ti.com/lit/ml/slyp167/slyp167.pdf

[Josh Perfetto]

Thanks for the link. What noise/dynamic response have you been able to
achieve with the open spectrometer?

-Josh

[Nathan McCorkle]

I'm still learning KiCAD, so I haven't even got that far yet. Any free
time to help?

[Josiah Zayner]

Here is also a decent reference I try and keep around http://zone.ni.com/devzone/cda/pub/p/id/262

I am sure there are lot of other things out there also.

Software can deal with noise alot also. Filtering algorithms are a large part of modern signal detection. Read up on signal processing. There are many theoretical ways to increase your resolution with just code.

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

[John Griessen]

On 05/14/13 23:39, Josh Perfetto wrote:
> But don't you have an issue with noise limiting your dynamic range if
> you just take a camera designed for 8 bits and put a 16 bit ADC on it?

A CCD is a component of a camera, not a camera, so putting a well matched
filter and ADC at the output of its bucket chain makes some sense.

> Like say your CCD holds 0-100,000 electrons. That's more than your 16
> bit ADC so the ADC is your limiting factor.

0 is too absolute of number to use here when talking in noise terms.

If you really mean it's a detector that can resolve dark vs one electron
all the time, it's the kind that lives in a Dewar flask of liquid helium
and not what we are usually talking about here.

But then you get 50
> electrons of noise in each pixel, so now your dynamic range drops to
> 2000, so you are effectively getting just an 11 bit camera? (Or even
> worse, just some example numbers)

"50 per pixel" says you're talking about two more different things.

A CCD bucket brigade circuit moves charge along with some noise added,
but not a ton, and never deals with single electrons, but with much
larger numbers of them.

Back to "take a camera designed for 8 bits and put a 16 bit ADC on it?"
You're probably meaning integrated detector/ADC chips, but a CCD is analog
like Nathan said, and it's a component system that you can tweak, unlike
a camera chip, so yes, you can put a higher res ADC on a CCD than it is usually
spec'd for. Not sure what it would buy you -- case by case -- read the,
(physics intensive), datasheets as usual.

[Josh Perfetto]

On Sat, May 18, 2013 at 3:23 PM, John Griessen <jo...@industromatic.com> wrote:
> Back to "take a camera designed for 8 bits and put a 16 bit ADC on it?"
> You're probably meaning integrated detector/ADC chips, but a CCD is analog
> like Nathan said, and it's a component system that you can tweak, unlike
> a camera chip, so yes, you can put a higher res ADC on a CCD than it is
> usually
> spec'd for. Not sure what it would buy you -- case by case -- read the,
> (physics intensive), datasheets as usual.

This is exactly the point I was trying to make. I don't think you can
just take a well-designed 8 bit camera, which as you've pointed out
contains multiple components, and replace the ADC with a 16 bit ADC,
and expect to get 2^16 dynamic range out of the resulting camera,
mainly because of noise. Software signal processing can help some as
has been pointed out, but there's also a lot of hardware
considerations that start to increase the cost.

That was the basis of my interest in what the minimal dynamic range
requirements would be. I realized some flaws in my earlier analysis of
it. The point Patrik made about this being the bits in the mantissa
suggests a way to tie the dynamic range of the camera to the
resolution of Ct measurements. So before I thought if I want .01
resolution in Ct measurements, and each cycle doubles the signal at
100% efficiency, this means I need to have a dynamic range of 100 for
each cycle, which 7 bits would be sufficient for. This is ok so far,
but would result in only a single cycle which results in a non-zero
and non-saturated fluorescence value. However to draw the
amplification curves used by most analysis algorithms, I need several
data points across a series of cycles. I am not exactly sure how to
quantify or solve this mathematically yet, but it seems like something
like 4-16x more range is required than the above calculation suggests.

-Josh

[matt harbowy]

One of the things that I think is being missed is too much of a reliance on off-the-shelf webcams. I'd really encourage everyone to "get back to basics".

The model piece that I've been using is the AMS linear array. http://www.mouser.com/ProductDetail/ams/TSL1402R/?qs=sKasJQfA%252bi5t6//qMqmpuA== It's a 16 dollar chip that if you spend some time actually playing with it, youll begin to realize that many of the arguments being made here are moot, or mootable.

By controlling the signalling to the chip, you can control the integration time- the amount of time the chip spends "counting photons". Similar (or exactly the same thing) to exposure on a camera, you control the mantissa at the upper end of the range by reducing the amount of time the capacitors have to store the charge, the electrons being released by photons. If you need to resolve at the lower end, you can increase the exposure time to capture more events. 

To the point of "physics intensive"- you waste more time dealing with people who "cant do math" than just doing the math. What we need to do is start breaking apart the jargon. It annoys me when people talk about 16 bit ADC on a 8 bit camera. There's no such thing as an 8 bit camera- there's just capacitors storing charge off a diode, analog doesnt have bits, thats why it is analog. If you're hooking the ADC up right, its the physics of the underlying chip, the diode and capacitor, that really matters: the choice of wavelength (green is usually better), the tolerance and noise and things like "dark current" that dictate how well a 8-, 10-, 24-, 1000- bit chip are going to read that charge.

256 px doesn't seem like a lot, but it's all about controlling not just things like slit width, but also integration time and averaging, as well as all of the components of Beers law. A cheap ADC (Arduino supplies 10 bit) can cover way more range if you have control over exposure time, and attempt to adjust or autorange as best as possible. It would seem that for DIY qPCR, sinking all of your money into pricey detectors is wasted if your optics are crappy, and that sample averaging and exposure control can squeeze whatever resolution you need provided that you have decent control over the optics of the system.

I think part of the problem I'm having engaging the DIY community is that there's a willingness to salvage parts but not as much attempt to document and share what people learn along the way. Taking apart $$$k equipment is fun and all, but it seems like we're losing the basic skills that were state of the art 20 years ago, like the power of what you can do with a single, crappy cheap ass photodiode or photomultiplier. Yes, "its been done", and it's not what the cool kids at Illumina or BioRad are doing, but often simple can force you to think of new ways of looking at the problem. There's no reason why these kind of experiments aren't open sourced and repeated to death so that everyone, no matter what their experience level, can begin to communicate and share with others.

lets teach the physics, the math, the basic science, and not just gloss over it. do, don't debate. Grab an arduino and a photodiode and capacitor and do the experiment. mess around!

-matt

[Josh Perfetto]

Hey Matt,

My upcoming monochromatic qPCR machine is photodiode based, and Nathan
has been doing a lot of work with CCD linear arrays for his open
spectrometer project. We're starting R&D on a multichannel
multichromatic (500-750nm) qPCR project though, which will likely
require a 2D 12-bit camera. I would love to know what is "back to
basics" on such a device, as I am not an EE, and the off the shelf
solutions are pricey.

-Josh

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups
> "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/df579cd1-5fd6-4dff-8c69-2ade142a1450%40googlegroups.com?hl=en.

[Ashley Heath]

Josh - do you have more information on the monochromatic qPCR machine ? I need one!

[Jeswin]

On 5/20/2013 2:33 AM, Josh Perfetto wrote:
> Hey Matt,
>
> My upcoming monochromatic qPCR machine is photodiode based, and Nathan

Monochromatic meaning you can only use one type of probe for your
experiments?

[Josh Perfetto]

Ashley, the plan is to provide a very low-cost unit, capable of single
channel detection in 16 200 uL PCR tubes. The machine will have fast
ramp rates of at least 3 C/s, and very friendly software accessible
via wifi/ethernet/touch screen. The small sample size will limit its
usefulness in some research applications, but I'm interested to see
what new applications we can devise for it by offering it at a super
affordable price point. I'll blog more about this soon. More poweful
devices can come later (what this thread was about).

Jeswin, what I mean is it could detect any green fluorophore, but
would be only single channel (i.e. no multiplexing). So if you want to
run multiple probes, you have to run them side-by-side in multiple
tubes.

-Josh

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> --- You received this message because you are subscribed to the Google
> Groups "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/519AA7F5.5060401%40gmail.com?hl=en.

[Jeswin]

On Mon, May 20, 2013 at 7:05 PM, Josh Perfetto <jo...@openpcr.org> wrote:
> Ashley, the plan is to provide a very low-cost unit, capable of single
> channel detection in 16 200 uL PCR tubes. The machine will have fast

Why not PCR plates like in reqular qPCR machine? Maybe your detection
method differs so you can't use plates?

>
> Jeswin, what I mean is it could detect any green fluorophore, but
> would be only single channel (i.e. no multiplexing). So if you want to
> run multiple probes, you have to run them side-by-side in multiple
> tubes.
>

I see. You can use different channels but can't have multiple channels
in one reaction.

[Josh Perfetto]

On Tue, May 21, 2013 at 8:58 AM, Jeswin <phill...@gmail.com> wrote:

On Mon, May 20, 2013 at 7:05 PM, Josh Perfetto <jo...@openpcr.org> wrote:
> Ashley, the plan is to provide a very low-cost unit, capable of single
> channel detection in 16 200 uL PCR tubes. The machine will have fast

Why not PCR plates like in reqular qPCR machine? Maybe your detection
method differs so you can't use plates?

Hi Jeswin,

Basically it was far cheaper to create this machine for 16 wells than for 96 wells, and I wanted to introduce something at a very affordable price range first. I think qPCR is a very powerful technique that too few people do because the hardware is so expensive. For many applications you can not only get quantitative data, but also avoid running gels which dramatically increases your workflow.

So at 16 wells it wouldn't be much of a plate, though of course the spacing is the same so you could cut up your 96 well plates into 8 16 well plates if you wanted. Or just use 8-tube strips.

-Josh

[Josiah Zayner]

People don't _not_ do Real Time PCR because the equipment is so expensive. They don't do Real Time PCR because it is not very reliable. Intra-experiment variability is very high. Protocols are complicated and all the reagents are expensive.

With RNA quality being a hugely important factor most home labs are not equipped to do anything successful.
Real Time PCR protocols that do more then detect copy number of a gene are complicated and require preventing DNA and RNAse contamination. You need to reverse transcribe your mRNA, chop up the DNA, chop up the RNA, purify the cDNA. And have multiple samples because your variability will be so high even on high-end machines much less a DIY machine.

With Sequencing/Deep Sequencing starting to become really cheap and you get to see the copy number of every transcript not just the ones you PCR it is becoming the goto technique.
The fact that one can just do Reverse Transcriptase PCR and run it on a gel if you want something quick and dirty makes Real Time PCR not a very good method. 

Real Time PCR is not a fancy technique that can do something nothing else can do.

Most Real Time PCR used nowadays is for diagnostic stuff.

Number of citations that contain the words "real time PCR" searched for on Google Scholar by Year

2012  47,400 (of these 23,900 include the word diagnostic)
2011  80,500
2010  104,000
2009  114,000 (of these 18,600 include the word diagnostic)
2008  113,000
2007  104,000
2006  85,100


There is a reason the citations have dropped off drastically as a research tool because people have found it is not a really good technique.

--

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2BL%3DET2ohCCVMULKPdouGQ5KSuB1TOYr0gB8G%3DZUriYcPtMejg%40mail.gmail.com?hl=en.

[Jonathan Cline]

On Monday, May 20, 2013 4:05:12 PM UTC-7, Josh W. Perfetto wrote:

Ashley, the plan is to provide a very low-cost unit, .... 

 very friendly software accessible
via wifi/ethernet/touch screen. 

Low-cost is not synonymous with providing these expensive system components (wifi/ethernet/touch screen).  Don't even mention using likewise expensive & underwhelming arduino.   $10 here, $10 there, it all adds up and is not very low cost.  USB is a better bet, serial still the cheapest and simplest. For an automation system, a better bet is industrial grade wired bus with longer distance.  Lab equipment typically doesn't move or alternatively if it's a field unit, it is plugged into a computer after logging data internally.

Here's typical options for low cost communication. It's probably better to have independent power rather than run power through the communication link.  

RS-232 - UART framed, single ended, full duplex, non shared media, NRZ, up to a few hundred feet, non powered, non isolated. 

RS423 - UART framed, single ended, full duplex, non shared media, logic level, several hundred feet, non powered, non isolated. 

RS-422 - UART framed, differential, full duplex, non shared media, logic level, up to a thousand feet, non powered, non isolated. 

RS485 - UART framed, differential, half duplex, shared media, up to a couple thousand feet, non powered, non isolated. 

CAN/Fieldbus - UART framed, differential, half duplex, NRZ, powered,  signal can be isolated.

GPIB/IEEE-488 - parallel bus, logic level, short-range, shared media, high bandwidth, ruggedized connectors, very common in tech equipment, more expensive connectors & cabling in comparison to the others.

[Nathan McCorkle]

Can you add the respective run lengths of each protocol?

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups
> "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/06c8c217-1410-4ce8-b09c-80d59b4cda88%40googlegroups.com?hl=en.

>
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

--
-Nathan

[Simon Quellen Field]

We may have different definitions of low cost.

:-)

You can get WiFi, a touch screen, and USB (which gives you Ethernet if you add an $8 dongle) all in a nice tablet package, for $56, including shipping.

Starting with that, adding a $3 microUSB to female USB connector, and writing some Android software, you can control a $10 TI MSP430 Launchpad (an Arduino on steroids) using the Android USB hosting feature.

So far, we're up to $70 to add WiFi and a touchscreen (and we also get local storage, a 1 GHz processor, programmability, and a much better user interface experience than a keypad and LCD display).

The devices you want to communicate with (laptops and the cloud) don't have RS232 serial ports. But they do have WiFi and USB, both of which are much faster.

I get my Internet here on the mountain using WiFi. My dish is aimed at a dish 4 miles away. So range is not an issue (and of course WiFi is isolated -- no wires). But once you have the Android tablet, range is not an issue anyway, since you are on the 'net, and anyone in the world can get to the data. The tablet can run an Apache server and serve the data itself.

A USB to RS232 adapter is more than $10, and is slow (most won't do more than 0.1 megabits). To do file transfers over it requires writing software, and adding error correction slows it down even more. With USB or WiFi, you just drag and drop the data, and it is error corrected. The Android tablet looks like a disk to your laptop, and looks like a web server to your browser. Controlling it and getting data to it and from it is a lot easier and faster using WiFi or USB than any RS232 solution.

Since the laptop already speaks USB and WiFi, why convert USB to slow serial, and lose all those advantages?

Of course the device doesn't need a touchscreen if it has WiFi and a Linux OS.

Use a $25 Raspberry Pi and a $10 WiFi dongle, and put the user interface in a web page on the Pi.

Now you control it with a laptop or a smartphone, using a web interface.

And you don't need another microprocessor, the Pi has GPIO pins accessible to control lights, motors, and servos.

If you want to save the $10, omit the WiFi dongle and just plug the Pi into your laptop's USB and set it up as a network device.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

On Wed, May 22, 2013 at 12:07 PM, Jonathan Cline <jnc...@gmail.com> wrote:

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/06c8c217-1410-4ce8-b09c-80d59b4cda88%40googlegroups.com?hl=en.

[Jonathan Cline]

Um, what's the ballpark price target then, for assorted devices under discussion? I suggest $30 total circuit cost as a target, not counting mechanical.   Don't forget the CCD if any, unless using  photodiodes.  Raspberry Pi is not too good for controlling I/O: can be seen by the projects adding on additional boards beyond the Pi.  "Low cost" seems to be frequently bandied about here, yet the result seems not very low cost at all.   And transferring very little bandwidth does not need expensive networking components.  Again, the lab devices don't move, so they can be connected to a nearby computer for the network connectivity/apps (labs have cheap computers, plenty of them), and the amount of data transferred is very small (kilobytes not megabytes, and speed is not required anyway). 

Ref: http://www.horizonpress.com/pcr/qPCR-machines.html
"""The following PCR machines are compared for various features to help you decide which instrument is most suitable for your needs. ... ABI StepOne, LightCycler 480, Mx4000, Rotor-Gene 6000, Mastercycler, MiniOpticon, MyiQm, Chromo4, iQ5"""

Sure, go add wireless to it.  Then be prepared to pay several thousand minimum to get it certified by the FCC for RF emissions.  Oh, it's OK, you'll just amortize that cost into the devices' selling price.  Wait, the idea was to be low cost, oops again.   And now it requires a $600 Android phone to be used?   Yikes, stop the insanity.
 

## Jonathan Cline
## jcl...@ieee.org
## Mobile: +1-805-617-0223
########################

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CAA0yOM60i1h7cf_1KEB_Lsc5zHGHACYHUmtER5LrtivYAHXV9w%40mail.gmail.com?hl=en.

[Josh Perfetto]

On Wed, May 22, 2013 at 4:46 PM, Simon Quellen Field <sfi...@scitoys.com> wrote:

We may have different definitions of low cost.

:-)

Yes I suppose "low cost" was too vague. I'm sorry that I haven't properly described this yet and will work on a more complete blog post shortly. My main goal is to bring what I think is a very powerful technique to a lot more people than are currently using it. Obstacles to doing that include hardware costs, reagent costs, and knowing what to do and how to do it for your application, and even developing new applications. I think that the touch screen is valuable for enabling proper use of the device/reaction setup in some use cases, and being a device which generates data, internet and web connectivity via ethernet/wifi are paramount to A) providing a great user experience for researchers to access and analyze their data, and B) enable others to build applications on top of the machine. By "low cost" hardware, I meant it would be cheaper than the cheapest generally available qPCR machine (which AFAIK is about $10k) by at least an order of magnitude, not that it would be the absolute cheapest machine possible.

Josiah earlier talked about one very important application of measuring gene expression. This is probably the biggest use of qPCR today and what originally got me into it (I wanted to "debug" biological systems I had built, and gene expression, while in no means adequate by itself, is one of the easiest and most effective tools). To the extent people have difficulty getting this to work, that is an opportunity for DIYbio to innovate. I am excited about this machine especially in a DIYbio context because unlike the earlier OpenPCR machine, this machine produces data, and there is a lot of things people on this mailing list can do with data once it is exposed in an open platform.

But there are applications well beyond that. Health apps are obvious: qPCR is the gold standard in viral diagnostics and quantifying viral load. But also say you want to know if there is horse meat in your food, or dolphin meat, or listeria, or E.Coli O157:H7? Again qPCR is the gold standard.

But I'm even interested in the most mundane uses. Say you have ran a PCR reaction, and now want to send it off for sequencing to do DNA barcoding. Or you PCRed something to clone. But you don't know if the PCR worked. Why should you spend time and money sending it off to sequence speculatively, only to find out days later that it didn't? Or why should you spend time and money casting and running a gel? If your PCR machine can show you the amplification curve and melting curve analysis right after the run, then you will immediately know if it worked. We need smarter machines like this and many others at every step of the way to make biology engineerable, because right now the whole process takes too damn long.

-Josh

[Jeswin]

>
> But there are applications well beyond that. Health apps are obvious: qPCR
> is the gold standard in viral diagnostics and quantifying viral load. But

Yep, that seems to be the most common uses for qPCR. Viral detection,
mutation assays. The company I work for, Reniguard, has been working
on these kind of assays.

> also say you want to know if there is horse meat in your food, or dolphin
> meat, or listeria, or E.Coli O157:H7? Again qPCR is the gold standard.
>

I thought about that sort of stuff, on the DIYbio level. But what
makes qPCR a better alternative than regular PCR and running a gel?
Getting cheap qPCR enzymes that also give good signals are difficult.
You can't use regular Taq.

> But I'm even interested in the most mundane uses. Say you have ran a PCR
> reaction, and now want to send it off for sequencing to do DNA barcoding. Or
> you PCRed something to clone. But you don't know if the PCR worked. Why

Colony screen them? In all of the cloning I have done, if the screen
is positive, then I am ~95% sure the sequence is fine. There was just
1 or 2 cases where there was an insignificant mutation outside the
region of interest.

[Jonathan Cline]

Touch screen: how's that work with gloves on? Better make it resistive.  Except, no one likes those because they're really annoying to use (pressure, precision), especially now that we're all spoiled by capacitive displays.  Worse case you've got a membrane keypad (again, no one likes those) and a non-touch display.  Either way you'll have to compromise on display resolution and size, when you really want to graph something onscreen in a really large format.   Meanwhile one of the most successful equipment stories is still Nanodrop, which doesn't have a display at all, as I suggested: it sends all data to the nearby computer. 

Re: Wireless again.  Measuring very low voltages with sensitive electronics while beaming a bunch of RF energy all around right next to the amplifiers will cause trouble.  

Cheaper than competitive equipment?  Why not kill off the competing equipment with either a retail price so low that their margins are destroyed, or alternatively keep the "higher low" price as you suggest and keep more margin yourself?   Either way the cost of building the device should be lower, not higher.  

On Wed, May 22, 2013 at 10:41 PM, Josh Perfetto <jo...@openpcr.org> wrote:

 I think that the touch screen is valuable for enabling proper use of the device/reaction setup in some use cases, and being a device which generates data, internet and web connectivity via ethernet/wifi are paramount to A) providing a great user experience for researchers to access and analyze their data, and B) enable others to build applications on top of the machine. By "low cost" hardware, I meant it would be cheaper than the cheapest generally available qPCR machine (which AFAIK is about $10k) by at least an order of magnitude, not that it would be the absolute cheapest machine possible.

--

[Josiah Zayner]

I agree with Jeswin almost completely.

No, offense but I don't think anyone will use a DIY machine for diagnostic uses and telling people that it can be used that way is just looking for a disaster. (I tested X on the machine and it said I have X disease....)

One can detect all those things you mention with normal PCR.

You want to see if you PCRed something and it worked. People run agarose gels for that. Doing Real Time PCR to try and figure out if your PCR worked really seems like overkill. Doing it instead of sequencing,  probably costs 10x in reagents and 1000x in time.

Debugging gene expression. A much easier way is protein expression or Reverse Transcriptase PCR as Jeswin said.

The data Real Time PCR generates cannot usually be compared between experiments. It is kind of like a western blot, there is way too much variability. Real Time PCR really has not much place in biological engineering. What can one use it for? There are numerous papers that show that gene expression has no correlation with protein expression. People run SDS-PAGE gels or Western Blots. The variability is why most people tend to not quantify these things. Instead just say, X band is obviously more than Y which is greater than Z.

In qPCR you run a control of a known "housekeeping" gene such a GAPDH or Actin to have a relative comparison. The transcription of these mRNAs is dependent on minutiae like state in cell cycle or time "alive" or nutrients or temperature or microenvironment. Things that cannot be replicated easily not matter how hard you try. It is very bad form to compare one run to another run, same thing for stuff like Reverse Transcriptase gels or Western blots. The variability in mRNA expression makes it so only relative comparisons are possible. Error propagation in exponential processes is huge. 

Again we have not even gone back to the handling of RNA in a home lab.

I have worked with RNA plenty doing Ribosome profiling and the lab I worked in is on lockdown from touching anything without a gloved hand, ethanol wash everything before use. RNAse Away everything and still people have RNA degradation problems. I don't know how much faith I could put in data generated from a home lab with very few resources and many sources of contamination.

I don't mean to be so negative and I hate telling someone not to do something without giving them other options. So...

• I think your time would be better spent developing assays and protocols that are easy and cheap but can answer big questions. Why not setup a DIY bacterial two-hybrid system so in vitro evolution is easily available to people?
• Why not develop software for use in bioinformatics?
• A cool program would be to use machine learning to analyze conformational changes from molecular dynamics simulations using output from GROMACS programs.
• How ions actually find channels so fast i.e. how ions overcome the 3D diffusion random walk problem?
• Micromanipulation of DNA using AC and DC current ala http://www.sciencedirect.com/science/article/pii/S0006349502753988
• A non sequence based protein structure classifier using things like ramachandran angles, hydrogen bonds &c.

If you want to work on hardware

• An easy, cheap, DIY setup to make a light microscope a decent fluorescence microscope.
  
• A modern AFM(I know there is bunch of old stuff floating around on these things don't know about new stuff haven't looked it up in a while)
• A diode array or scanning spectrophotometer like what Nathan is trying to do. Maybe one that costs < $50 and has nice software and 2 nm or better resolution.

There are so many cool things and ideas to work on it seems like a waste of effort to put it into a Real Time PCR machine that wouldn't be extremely useful but in the end you should do what excites you the most even if other people think it is a bad idea.

--

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---

You received this message because you are subscribed to a topic in the Google Groups "DIYbio" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/diybio/Grxk60iX4Uk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to diybio+un...@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CAAhtNQs5rtMYwvxyMBztKa%2BJLgk%2BVfNtDMJSZy1pDabhR5JeoA%40mail.gmail.com?hl=en.

[Josh Perfetto]

On Thu, May 23, 2013 at 6:03 AM, Jeswin <phill...@gmail.com> wrote:

> also say you want to know if there is horse meat in your food, or dolphin

> meat, or listeria, or E.Coli O157:H7? Again qPCR is the gold standard.

I thought about that sort of stuff, on the DIYbio level. But what
makes qPCR a better alternative than regular PCR and running a gel?
Getting cheap qPCR enzymes that also give good signals are difficult.
You can't use regular Taq.

Well for something like GMO detection, you probably want to quantify how much there is. For something like Listeria, maybe you don't care exactly how much there is, but only want to know if there is any at all. Then yes you could do endpoint PCR and run a gel. If you already had a PCR machine and gel apparatus this might be the easiest way to run one sample. But there are some advantages to the qPCR:

- If you already have a qPCR machine, I think qPCR will be cheaper than casting and running a gel, and will *certainly* save a lot of time. There are cheaper enzymes out there, and also dyes like EvaGreeen.

- If you are doing any sort of real testing with lots of samples, you may not want to open the tubes after PCR. During the course of related testing for this machine, I ran 100's of PCR reactions and was doing many things with amplicon products like putting them in fluorometers and running gels to compare. I was not overly careful about this and now my negative controls often amplify for this assay, even with different pipettes. And I did check that it is amplicon and not phage contamination as when I switched the primer set to another region, the negative controls were ok again.

One other way to do this cheaply in a DIYbio context: prepare your PCR reaction, and then aliquot it into 2 tubes. Run only one in PCR. Then add a cheap dye like GelGreen to each, and compare the tubes in a transilluminator, and look for an obvious increase in fluorescence in the PCR tube. Of course you won't know it's the correct amplicon, but this is an option.

> But I'm even interested in the most mundane uses. Say you have ran a PCR
> reaction, and now want to send it off for sequencing to do DNA barcoding. Or
> you PCRed something to clone. But you don't know if the PCR worked. Why

Colony screen them? In all of the cloning I have done, if the screen
is positive, then I am ~95% sure the sequence is fine. There was just
1 or 2 cases where there was an insignificant mutation outside the
region of interest.

My point was to not even colony screen them unless you knew the amplification was good. This is more of an engineering point that I don't think is very commonly held by biologists right now, but should be. Basically I find myself often doing one step to verify a previous step, or doing a multi step process with no intermediate verification, and then realizing a failure occurred only at a later step. So I not only wasted time in doing later steps, but have to spend more time finding where the failure occurred. If verification happened at each step, especially if it happened cheaply, transparently, and automatically, that would be a big win for engineering biological systems.

I first did qPCR because I wanted to measure gene expression, but then after seeing this, I never wanted to go back to endpoint PCR, because of this benefit. Once you have the machine it is cheap, so I see no reason not to do it, unless you have a specific case where fluorophores will not be cleaned up and will cause a problem.

-Josh

[Josh Perfetto]

On Thu, May 23, 2013 at 11:54 AM, Jonathan Cline <jcl...@ieee.org> wrote:

Touch screen: how's that work with gloves on? Better make it resistive.  Except, no one likes those because they're really annoying to use (pressure, precision), especially now that we're all spoiled by capacitive displays.  Worse case you've got a membrane keypad (again, no one likes those) and a non-

Doesn't the iPhone/iPad/Android phone work ok for you with gloves on? I agree anything less than capacitive is not desirable.

 

touch display.  Either way you'll have to compromise on display resolution and size, when you really want to graph something onscreen in a really large format.   Meanwhile one of the most successful equipment stories is still Nanodrop, which doesn't have a display at all, as I suggested: it sends all data to the nearby computer. 

Yes, this is why I wanted to have really good IP connectivity so it could be controlled from any computer, and data analyzed in a convenient way. There were other use cases I wanted to support though, especially someone doing the same assay continuously, like in a food testing application. It may be easier to select the assay, enter a sample identifier, and hit go in the lab, without fussing with a computer.

Re: Wireless again.  Measuring very low voltages with sensitive electronics while beaming a bunch of RF energy all around right next to the amplifiers will cause trouble.  

That is a really good point, I am glad you raised it. I had not thought about that or tested it specifically. I will try to see what impact it has but I think you are right. Perhaps the radio can be depowered when measurements are taken.

 

Cheaper than competitive equipment?  Why not kill off the competing equipment with either a retail price so low that their margins are destroyed, or alternatively keep the "higher low" price as you suggest and keep more margin yourself?   Either way the cost of building the device should be lower, not higher.  

Well that is kindof what I think being an order of magnitude cheaper does, except that I am not trying to compete directly with the existing equipment. The specs of this device (16 wells, single channel) certainly don't match most existing equipment, so it will not be a substitute for many existing users. But I think there are many people who currently don't do qPCR, who could use this machine despite those limitations, if the price point was much lower.

-Josh

[Nathan McCorkle]

I think having even a single-tube qPCR machine coupled with some
gelRed or gelGreen would be better than running a gel, because the dye
isn't wasted when you load the sample into a gel well... so you can
see real-time if your reaction is doing ANYTHING AT ALL, then see what
the band sizes are, etc later on the gel.

Sounds like the variability Josiah is talking about is in sample prep,
so just produce the machine at a lower cost and it will still appeal
to the people with the variable protocols that can still pull grant
money.

Josh, simply add a capillary tube in same box with a high-voltage
power supply, and now you've got a reagent-reduced capillary
electropohoresis AND a sensitive optical path in place for detecting
fragment elution time. Now you're setup to increase the capillary
length, and find/formulate cheaper non-agarose gel, and you've got a
sanger sequencer. (Actually I found a paper that acquired 2bp
resolution with agarose, but it wasn't amazing....)

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups
> "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/CA%2BL%3DET29bDH36EnFb-aJ%2B4rhRKcdeVK1GmR%3Du2XhrLKUOPMtAw%40mail.gmail.com?hl=en.

>
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

--
-Nathan

[Josh Perfetto]

Hi Josiah,

Obviously a lot of regulation happens after transcription, but that doesn't mean that measuring mRNA is unimportant. That's why many people do it, and there are many useful biomarkers from RNA alone. For quantification, you can also do absolute quantification and so not depend on any "housekeeping" genes. You can also compare data between runs as long as you do an interplate calibration. You may also be able to design your experiment to avoid the need for this, like if you put all assays for the same gene in the same run.

Just like OpenPCR wasn't only for DIYbio, this won't be only for DIYbio either. But I don't like putting limits on what DIYbio can do like saying you can never handle RNA in a home lab. I think people will surprise you.

-Josh

You received this message because you are subscribed to the Google Groups "DIYbio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CAEUkM4sUPj82Hs9KG2cQGnJoMa4Dgs6Zg0TxQUV7vZ%2BixbDrLg%40mail.gmail.com?hl=en.

[Josh Perfetto]

On Thu, May 23, 2013 at 2:00 PM, Nathan McCorkle <nmz...@gmail.com> wrote:

Josh, simply add a capillary tube in same box with a high-voltage
power supply, and now you've got a reagent-reduced capillary
electropohoresis AND a sensitive optical path in place for detecting
fragment elution time. Now you're setup to increase the capillary
length, and find/formulate cheaper non-agarose gel, and you've got a
sanger sequencer. (Actually I found a paper that acquired 2bp
resolution with agarose, but it wasn't amazing....)

 Yes, maybe for version 3 :)

-Josh

[Cory Tobin]

> I think having even a single-tube qPCR machine coupled with some
> gelRed or gelGreen would be better than running a gel,

You'll definitely need more than one tube. You have to run controls
simultaneously to the experiments because experimental qPCR results by
themselves are meaningless and you can't directly compare data between
runs. And I would recommend doing every control and experiment in
triplicate. So if you wanted to test how a single gene changes
expression level after a certain perturbation, you would need need 3
experimental tubes and 3 control tubes before the perturbation and 3
experimental and 3 controls after the perturbation. Additionally,
unless you're working with a well established protocol like a viral
titer assay, I would recommend using multiple dilutions of your
controls.

A long time ago I was working on the ENCODE project trying to confirm
ChIP-Seq and RNA-Seq data coming out of an experimental Solexa
sequencer using qPCR. My general rule of thumb was that I needed 12
tubes to do one test if I was using a 96 well plate and 18 tubes if I
was using a 384 well plate because the smaller volume causes larger
pipetting error. So I would say the absolute minimum amount of tubes
you need for a qPCR machine would be 12.

-cory

[Nathan McCorkle]

Reading a CCD datasheet just now, it defines dynamic range (DR) as
DR=Vsat/Vmdk... where Vsat is the saturation output voltage, and Vmdk
is the maximum dark noise voltage

> --
> -- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.

> To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CAC46Jss%3DaiJAJouNKH02CjYP7NLemYN0QBdgW%2Bp-ymXmhRZuww%40mail.gmail.com?hl=en.

[Jesse vanWestrienen]

Hello all, first time post.  Thought I would bring a couple of projects to your attention. 

First is out of UC Berkley 

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0070266 

The cool part is they provide in a zip file all of the schematics and source code etc.  This may help answer some of the questions being thrown around here.  

The second is a project I am involved in, we have built a low cost qPCR machine. 

www.bio-meme.com

Josh cant wait to see yours in action. 

Jesse 

On Monday, May 13, 2013 4:04:47 PM UTC-4, Josh W. Perfetto wrote:

Hello,

Does anyone have a sense of what real time PCR instruments typically offer for the dynamic range and resolution of the fluorescence detection (and NOT the dynamic range of the overall detection)? I am trying to determine the minimal requirements for a cheaper machine.

-Josh

[Jesse vanWestrienen]

Here is the zip file from the plos one paper, 

Jesse 

[Harriet Weatherford]

Nitro-memantine is not available yet as an agent to reverse Alzheimer's. see

http://beaker.sanfordburnham.org/2013/06/reversing-the-loss-of-brain-connections-in-alzheimers-disease/

Since there are so many waiting for a medication like this for our loved
ones, and it appears to be a combination of nemenda and nitro glycerine
I challenge our group to figure out a DIY so doctors could prescribe the
two drugs independently and this group could provide a road map on how
to combine the two drugs. I know there is the combined talent in this
group to do this. Thank you ahead of time.

[Nathan McCorkle]

A lot of literature mentions "nitromemantines", note the S on the end,
this implies there are more than one nitromemantine molecule. The
specific molecule is probably important to the efficacy as a drug.

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> --- You received this message because you are subscribed to the Google
> Groups "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.

> Visit this group at http://groups.google.com/group/diybio.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/522DEF7E.4060205%40yahoo.com.