Dear Stephen

Thank you for your very helpful answer. I also have some extractions that have a much lower yield (approx. 5 - 15ng/ul) as measured by Qubit, again from 150-200mg starting material. Would these samples still be ok for a microbiome analysis?

Thank you again.

Philip

Thank you Stephen. I was thinking that as my extractions are total gDNA, that it would be hard to estimate what percentage would actually be microbial DNA, but I think that from your answer I have more than enough!

Thank you - In relation to DNA quality again, another question. I have had a problem using bead beating for my DNA extractions.. There is nealry always a very faint band, with a smear on the gel. As I have said above, I would get Qubit readings of 50-300ng/ul from these same samples. My intention is to run 16S PCR and Pyrosequencing on these samples. Do you think that if these are the readings I am getting on Qubit, that it the DNA is of good enough quality for 16S PCR - pyrosequencing? I know that there is enough DNA there, and I think that Qubit is very sensitive in measuring dsDNA, is this a good measure of quality however? Thank you

I should mention that the bead beating is just a pre-treatment before the gDNA is extracted using the Qiagen Stool DNA mini Kit

Good suggestions by all. The only thing I would add, based on extracting and running hundreds of 16S samples through pyrosequencing in our lab each month, is that DNA yield (more specifically, conc. and quality) from the extractions are not perfect predictors of how the samples will perform. Checking the strength and quality of the gel band after tags and linkers are added is about the best predictor we have found.

Thank you Stephen. I have since received some gel results from a subset of samples sent for pyrosequencing last year, and have bands of expected length (approx. 300bp) for each sample. These are a little fainter than the test PCR, as the seq company only ran 5 cycles on the PCR integrating the adapters, but are very clear. In your experience, that is a decent predictor of how they'll perform on the sequencer (MiSeq in this case)?

Hi Philip for all of our microbiome samples we see a big disconnect between the DNA concentration measured by Nanodrop vs Qubit. Qubit is typically 2-4x times lower. but it perfectly correlates with qPCR data. So I think in many cases when isolating DNA from the challenging samples like stool and soil- there is a substantial amount of inhibitors co-purifying with DNA, that contribute to OD. and you might think you have a high concentration of DNA, while its mostly inhibitors absorbing at the same wave length. Qubit assays are DNA-specific and seem to work much better for microbiome samples- vs just OD reading

Hi Alexander, thanks for your input. Yes I now measure first with nanodrop, and then with Qubit and obtain readings between 20 and 300ng/ul usually. Are these yields around the same as what you are seeing with your samples?

Hi Philip we just launched a next gen PureLink microbiome DNA purification kit- that enables fast isolation of high-quality microbial (and host, where applicable) DNA from a wide variety of sample types, including challenging samples such as stool and soil - in addition to urine, saliva, misc swabs, transport media, growth media. As its doing a good job at inhibitor removal, now Nanodrop and Qubit readings are very close. For 200 mg input you generate a substantial amount of DNA, more than enough for sequencing. eg for human stool typically 20 ug.

http://www.thermofisher.com/order/catalog/product/A29790?ICID=search-product

http://www.thermofisher.com/us/en/home/life-science/dna-rna-purification-analysis/dna-extraction/microbiome-dna-extraction.html