7 replies to this topic

[maruca8]

Greetings! We need to perform a large number of PCRs in DNA samples that probably will be have a negative result (18S Plasmodium) .  
We also want to perform an intern quality control of the quality of the DNA by performing a Betaglobin PCR in these samples (Human B-globin gen)?

Is there a way in which we can save money? For example, can we pool the DNA of X (20-50) number of samples and  then perform the PCR?

Thanks in advance,

Maria Eugenia

[Trof]

I don't understand it. You want to detect Plasmodium in your samples? How can you pool them, then you won't know which sample is positive and which is not? That doesn't make sense.

You can theoretically save some money running a duplex reaction, both 18S and beta-globin in one well. But this is only possible if you're using Taqman format (fluorescent probes), because you need two probes with different fluorescent tag. But Taqman is more costly than SYBR reaction. If you calculate it, maybe you'll find, that it actually don't save money. It depends on how much samples you have, prizes of the probes, duplex optimization, and so on.

You can also save money using smaller reaction volume (10 ul instead of 20 if your cycler support it) but that is more prone to pipetting error and evaporation.

[Micro]

Hi Maria,

One way to pool samples for PCR, but still be able to identify which sample is positive is to.... how to explain this... imagine a 96 well plate with each well containg a different DNA sample. For each horizontal row take a sample from each well (i.e. A1 +A2 + A3 etc.) and pool them into a PCR tube. Then, for each vertical column take a sample from each well (i.e. A1 + B1 + C1 etc.) and pool them into a PCR tube. At the end of this process there will be 20 PCR reactions that represent 96 samples. If you get a positive your gel will have 2 positive bands... one in a rows PCR tube and one in column PCR tube, you use this like grid positions to identify which well was positive (i.e. row B, column 3 so you know B3 is positive).

Things you will need to check/consider before using this method:
- your PCR will still detect a positive DNA sample once you have diluted your DNA during pooling of the samples
- your technique and handling of the samples is accurate enough to prevent cross contamination between wells

Let me know if my explaination isn't clear. Good luck with your samples!
Cheers
M

[gebirgsziege]

With pooling you always run the risk to dilute and underestimate samples; and if one of your pooled samples is positive you will have to double check it to be sure (even if you are using the elegant method suggested by micro). The positive control must be done seperately anyway, as you never can be sure if not one of your samples in the pool is not working, as the other samples in the pool are positive.

[maruca8]

Trof, on 26 July 2010 - 12:44 AM, said:

I don't understand it. You want to detect Plasmodium in your samples? How can you pool them, then you won't know which sample is positive and which is not? That doesn't make sense.

You can theoretically save some money running a duplex reaction, both 18S and beta-globin in one well. But this is only possible if you're using Taqman format (fluorescent probes), because you need two probes with different fluorescent tag. But Taqman is more costly than SYBR reaction. If you calculate it, maybe you'll find, that it actually don't save money. It depends on how much samples you have, prizes of the probes, duplex optimization, and so on.

You can also save money using smaller reaction volume (10 ul instead of 20 if your cycler support it) but that is more prone to pipetting error and evaporation.

Hi! Thanks for your response. Maybe my question was not clear. We have blood samples that on microscopy were negative to Plasmodium sp. But we still want to confirm their negativity by PCR.

[maruca8]

Micro, on 26 July 2010 - 09:56 PM, said:

Hi Maria,

One way to pool samples for PCR, but still be able to identify which sample is positive is to.... how to explain this... imagine a 96 well plate with each well containg a different DNA sample. For each horizontal row take a sample from each well (i.e. A1 +A2 + A3 etc.) and pool them into a PCR tube. Then, for each vertical column take a sample from each well (i.e. A1 + B1 + C1 etc.) and pool them into a PCR tube. At the end of this process there will be 20 PCR reactions that represent 96 samples. If you get a positive your gel will have 2 positive bands... one in a rows PCR tube and one in column PCR tube, you use this like grid positions to identify which well was positive (i.e. row B, column 3 so you know B3 is positive).

Things you will need to check/consider before using this method:
- your PCR will still detect a positive DNA sample once you have diluted your DNA during pooling of the samples
- your technique and handling of the samples is accurate enough to prevent cross contamination between wells

Let me know if my explaination isn't clear. Good luck with your samples!
Cheers
M

Thanks for the explanation, it is quite clear the way in your mix the samples (and I must say quite clever!). I have just another question, I will extract the DNA, but then I suppose I have to measure the quantity of the DNA for each sample, and after that I will mix equal quantity of DNA? After that, I will always take 1-2 ul of the mixed DNA?Is that the way to proceed? Or do I only dilute 1:8 my said 8 samples and then mix them?
Thanks so much to all!
Maria

@gebirgsziege: Thanks for the advice, definitely we will have the positive control done separately. I am thinking of performing first some tests in known positive and negative samples to check that the PCR is able to detect them accurately. Thanks for the advice, it is quite appreciated!!

Edited by maruca8, 27 July 2010 - 02:05 PM.

[Micro]

How to pool the DNA ultimately depends on your samples and your PCR. I use this method with metagenomic work which involve cosmids libraries. For this type of work I add 1uL of each DNA type (approximetly 100ng/ul) into a tube and then add 1uL of the DNA mix to a 25uL reaction. This works out to between 8 - 12 ng of each sample DNA per reaction, which is enough for my work.

Given that you are looking to detect plasmodium that are at best weak positives in blood (since they are negative through microscopy) the first thing is to find out what is the lowest amount of DNA you can add before a weakly positive control will produce a false negative? This will give you an indication of how many ng/ul you need to add, and then you can work from there to determine your method of pooling your DNA.

To add to Gebirgsziege's comment about positive controls. With a project that sound like diagnostic sampling having a weak positive control as well as a regular positive control is a good idea. A weak control that is close to the lower limit of detection will show if there is a problem with your reaction mix that could prevent weakly positive samples from being detected.

[phage434]

Note that none of these "mix the sample" techniques help you if you need to establish good DNA isolation and presence of control genes in your samples.  You won't be able to tell if some of your samples in the mix fail to have any DNA, or if they have poorly prepared or contaminated DNA.