SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   Metagenomics (http://seqanswers.com/forums/forumdisplay.php?f=29)
-   -   Illumina 16S libs - tech & QC questions about construction and data comparison (http://seqanswers.com/forums/showthread.php?t=41644)

capsicum 03-13-2014 09:10 PM

Illumina 16S libs - tech & QC questions about construction and data comparison
 
I wanted to ask for people's opinions and experiences:

I'm looking at making some Illumina-compatible sequencing libraries. Illumina has a two-step PCR method where the first PCR amplifies a region of the 16S gene and the second PCR adds adapters.

1. Does anyone carry out the first and second PCR steps together on the one reaction? I read about someone doing this, but I can't seem to find it again.

2. What QC do you do on your primary and secondary PCRs? How do you deem a success or failure? Just band visibility on a gel?

3. How do you deal with a project where some samples show very strong amplification and others showed only just sufficient amplification for sequencing? Do you optimise the PCR (by diluting the template, for example) to increase amplification of the poorly performing samples, or just sequence them anyway if there's enough product to do so?

4. If you dilute your samples to obtain better amplification, do you think this needs to be done for all samples that will be compared? I.e. do you think it will bias the results if you sequence samples that have been treated differently in this way and then compare them? If not, then do you think it's legitimate to compare samples that have radically different amplification efficiencies?

5. Do you think it's important to normalise the input DNA concentrations for samples that will be compared? What about the input source material? We have some projects where samples that need to be compared might contain DNA concentrations that differ by 100x or more.

DRYTCYV 03-14-2014 01:32 AM

Quote:

Originally Posted by capsicum (Post 135041)
1. Does anyone carry out the first and second PCR steps together on the one reaction? I read about someone doing this, but I can't seem to find it again.

- The 2nd PCR reaction efficiency is influenced from the 1st reaction,since the insertion of adaptors depend of a specific sequence from the 1st step


Quote:

Originally Posted by capsicum (Post 135041)
2. What QC do you do on your primary and secondary PCRs? How do you deem a success or failure? Just band visibility on a gel?

- Qubit and gel (1st),
- Qubit, gel and qPCR (2nd) to normalize.

Quote:

Originally Posted by capsicum (Post 135041)
3. How do you deal with a project where some samples show very strong amplification and others showed only just sufficient amplification for sequencing? Do you optimise the PCR (by diluting the template, for example) to increase amplification of the poorly performing samples, or just sequence them anyway if there's enough product to do so?

- Be sure that the sample quality is the same, run a gel (1%) before you start and check the integrity. And of course load the same dna amount (measured by Qubit)

Quote:

Originally Posted by capsicum (Post 135041)
4. If you dilute your samples to obtain better amplification, do you think this needs to be done for all samples that will be compared? I.e. do you think it will bias the results if you sequence samples that have been treated differently in this way and then compare them? If not, then do you think it's legitimate to compare samples that have radically different amplification efficiencies?


5. Do you think it's important to normalise the input DNA concentrations for samples that will be compared? What about the input source material? We have some projects where samples that need to be compared might contain DNA concentrations that differ by 100x or more.

- The protocol information for the 1st pcr its 2,5 ul of a 5ng/ul sample. So if you want to compare samples the best thing its to proceed the same way. Load the same, in the begining, ->clean-up->gel -> 2nd PCR ->clean-up Qubit+gel+qpcr and normalize

I hope these will help you

capsicum 03-16-2014 02:23 PM

Quote:

Originally Posted by DRYTCYV (Post 135051)
- The 2nd PCR reaction efficiency is influenced from the 1st reaction,since the insertion of adaptors depend of a specific sequence from the 1st step

Yes, but how reliable do you normally find them? Some people must find them reliable enough to perform both at the same time (in the same reaction). I'd be worried that the product from the 1st PCR was too little. I guess you'd see that reflected in the final product (or lack of it) anyway.

Quote:

Originally Posted by DRYTCYV (Post 135051)
- Qubit and gel (1st),
- Qubit, gel and qPCR (2nd) to normalize.

This is what we'd planned to do too, except for the gel... many of our samples are far to low in concentration to visualise on a gel.


Quote:

Originally Posted by DRYTCYV (Post 135051)
- Be sure that the sample quality is the same, run a gel (1%) before you start and check the integrity. And of course load the same dna amount (measured by Qubit)

We can't always be sure of the quality because the DNA comes from many different sources (but all samples to be compared to each other) and usually the DNA is very low concentration (just a few ng/uL)

Quote:

Originally Posted by DRYTCYV (Post 135051)
- The protocol information for the 1st pcr its 2,5 ul of a 5ng/ul sample. So if you want to compare samples the best thing its to proceed the same way. Load the same, in the begining, ->clean-up->gel -> 2nd PCR ->clean-up Qubit+gel+qpcr and normalize

I know Illumina has suggested that mass and volume, but does it really make a difference? Theoretically, do you think it is more scientifically sound to use the same DNA mass or the same total original sample mass/volume before DNA extraction? I don't think it matters for the purposes of comparison... perhaps only for technical reasons in order to get the PCR to amplify efficiently. Often we have to dilute our samples before they can be amplified, even when the DNA concentration is low (lots of inhibitors that are very hard to remove, and very low cell content to begin with).

DRYTCYV 03-19-2014 03:23 PM

Quote:

Originally Posted by capsicum (Post 135184)
Yes, but how reliable do you normally find them? Some people must find them reliable enough to perform both at the same time (in the same reaction). I'd be worried that the product from the 1st PCR was too little. I guess you'd see that reflected in the final product (or lack of it) anyway.

Maybe they use a huge amount of DNA and like that its possible to perform both things at same time. So if you process eveything in the same pcr step, you will get indexed fragments and since you load more its possible to continue to sequencing step.


Quote:

Originally Posted by capsicum (Post 135184)
This is what we'd planned to do too, except for the gel... many of our samples are far to low in concentration to visualise on a gel.

Bioanalyzer?



Quote:

Originally Posted by capsicum (Post 135184)
I know Illumina has suggested that mass and volume, but does it really make a difference? Theoretically, do you think it is more scientifically sound to use the same DNA mass or the same total original sample mass/volume before DNA extraction? I don't think it matters for the purposes of comparison... perhaps only for technical reasons in order to get the PCR to amplify efficiently. Often we have to dilute our samples before they can be amplified, even when the DNA concentration is low (lots of inhibitors that are very hard to remove, and very low cell content to begin with).

No, i think you can play around with the starting sample quantity. I just said to load the same amount because its more easy to detect some problems between samples (like contaminants). Do you have enough sample to try to purifiy? Precipitation/purification kits? Your efficiency problem could be correlated with some contaminat in the solution...


All times are GMT -8. The time now is 05:30 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.