300 Stool Samples - 16s rDNA amplification & sequencing for identifying bacteria

[vs92]

I am trying to plan for a large experiment of 300 human stool samples over the next 1 month to identify the population of bacterial species living in each stool sample. Depending on the success of the experiment, I need to repeat ~ 300 human stool samples every month for the next few months. I was wondering if 454 was the right machine for this application?

[Dynamac]

I'm not familiar with the Illumina, but this will be quite an undertaking for the 454, since you have 300 separate samples that you presumably cannot pool. You'll need to investigate the use of MIDs on a 16-region PTP. With 16S, you can get away with unidirectional sequencing, which will help.

Once you have the machine and all the protocols, you'll be able to do the cost analysis.

[ajthomas]

This project is totally doable on a 454. You can run that many samples in one run on a FLX, although you might want to do it in two runs to increase your sequencing depth. A Jr. will need about 8 runs to equal the sequencing output of one FLX run, so that's still doable in a month.

As for cost, it's not worth buying a FLX (~$500K) for this project, but it might be worth buying a Jr. (~$100K) You might also consider just doing your sample prep and sending it out to a center for sequencing. They could do it all in one run, then if you decide that you need more data, you can have it run again and add the data from the second run to the first.

The reagent list price for a FLX run is about $7000. I've never seen official prices for Jr. reagents, but I think it's about $1000 or a little more. Actual prices will likely be less depending on your quote.

Regarding barcodes (MIDs), if your amplicon is only 300 bp long, you can dramatically reduce the number of MIDs(and therefore primers) you need by mixing MIDson both ends of the amplicon. For example, sample 1 has MID1 on the 5' end and MID1 on the 3' end, sample 2 has MID1 on the 5' end and MID2 on the 3' end, etc. You could do the whole thing with only 12-13 of each primer instead of 150 of each (assuming you use a 2-region plate on a FLX), which will save you a lot of cost when you buy your primers. If you use a Jr or smaller regions on a FLX plate, you will need even fewer primers.

There are several online tools for analysis of 16S data from 454 sequencing. I've only looked at them a little and I don't use them, so I can't really point you to anything specific; sorry. From what I have seen, and from what I've learned talking to colleagues who do use them, you basically just upload your data and wait for the analysis to come back, which takes a few minutes to a few hours. At least some of these tools BLAST the results as well and give you an identification for each sequence.

[vs92]

Thanks so much for the helpful information. I'll read more about how 454 can be used. From talking to colleagues, folks seem to recommend the MiSeq from Illumina for this application. I need to do a lot more reading and then choose from these two options....

[Palecomic]

Take a look at QIIME http://www.qiime.org and Mothur http://www.mothur.org for the analysis pipeline. BLAST isn't good enough for identifying these sequences, it does local alignments not global, but both pipelines have several methods of identification.

Also, 300 a month very doable, the limiting step will become how fast you can do the DNA extractions and make the amplicons. If you're doing this regularly I think a large set of barcoded primers is worth investing in, you don't have to do as much thinking on the pipeline side (we have a set of 576 we use from the Knight group) http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2584711/. They last a good while in aliquots in 96 well plates.

[elizondas]

454

Dear ajthomas,
I would like you help me to plan my study. We want sequence lot of samples, over 450, to identify the bacterial population to spcies level living in the nasopharinx before and after a flu infection and at differents times, day -1, 0, 2, 3, 6, 28 . I was wondering if 454 was the right machine for this application and how to design the primers in this approach. I have never done sequencing before.