Pool amplicons on a plate

[tng012]

I am new to pyrosequencing. I am working on 16s rRNA from environmental samples. I want to measure species diversity, I sequence my fragments from one end of the DNA molecule using the 454 amplicon titanium method.
One topic in the forum: "Using multiple MIDs in Titanium sequence runs" showed that we can use 10 MIDs-tagged libraries (shotgun samples) in 1 region, but for amplicon sequencing, how many PCR amplicons with diffirent MID-tagged can I pool in 1 region? Can I pool 10 MIDs-tagged in 1 region?

[kmcarr]

Theoretically the number you can pool would be huge. The question isn't really how many can you run in one region, it's how many should you. The practical limit is mostly due to how many reads you need to obtain from each sample. According to the Roche/454 literature a single large region should produce 360K - 520K high quality amplicon reads (reality is closer to the low end number). How many reads do need to get from each environmental sample to produce an accurate picture of its biodiversity? Divide that number into 360,000 to determine the maximum number of samples you should combine in one large picotiter plate region.

A very useful resource for microbial 16S rRNA information, and more specifically 454 amplicon sequencing of 16S rRNA can be found through the Ribosomal Database Project. They have a Pyrosequencing Pipeline set up for processing large 16S amplicon data sets obtained from 454 sequencing. Pay special attention to the help page in which they discuss the design of multiplex ID tagged primers. (Note that this page describes primer design for FLX Standard amplicons, not FLX Titanium, but the principles are the same.)

[flxlex]

You could also contact your local Norwegian 454 sequencing center, they might have experience with these kinds of projects. By the way, that would be me... ;-)

[martinjf]

Using regular MIDs "constraints" the number of multiplex to around 155 if I correctly remember it. An other option would be using tags combination in PCR primers. The system we developed allows for 5184 amplicons to be multiplexed. In practice we use around 1152 amplicons in 1/8 PTP plates. The only limit is to make sure that both tags are read during the sequencing as they are both needed for sample assignation. PCR products up to 300bp - 350bp seem fine so far (you can check that with simulating 454 flows on your model sequence).

The original protocol we developed in the lab is published here :

http://www.biomedcentral.com/1471-2164/11/296

We now use extensions of this protocol with longer tags, allowing for more safety in samples bioinformatics assignation.

Good luck !

Jef.