best 16s region to amplify with NGS ?

dph · 12-07-2011, 07:28 AM

Does someone have an idea of the best 16s gene region to amplify with NGS in order to assess microbial (bacterial) diversity in environmental samples ?
cheers

dph · 12-08-2011, 10:44 AM

no reply ??? no idea ?

themerlin · 12-09-2011, 06:03 AM

There has actually been a fair bit of work done on this in the literature. I'd look at some papers out of Rob Knight's group.

cliffbeall · 12-14-2011, 07:29 AM

One thing you might want to think about is doing more than one region. We looked at V1-2 and V4 on oral bacteria (pre-titanium) and found that V4 significantly undercounted the Bacteroidetes phylum, despite that the primers match Bacteroidetes perfectly well.

The Human Microbiome Project followed a similar tack, with their V1-3 and V3-5 primers. Our current strategy is to use their sequences, on the thought they will become a standard (at least for a few months until the technology changes again).

Jean · 12-19-2011, 09:58 AM

What environment? What are you trying to find or distinguish? What platform will you be sequencing with? The "best" region is the one that will work for the questions you want to answer. If, for example, you are trying to distinguish two species then one variable region might have more distinguishing power than another.

jrvalverde · 04-20-2012, 04:32 AM

Read the papers.

As already pointed out, it depends on what you want.

Certain regions work better than others to distinguish among different species, but cannot resolve among others. The best way to know is to find out studies about your target community and see whether they did work. For instance, v6 will not distinguish E coli from Salmonella, which is serious if you want to analyze health data, but may be irrelevant for soil data.

A second factor is the technology: if you are going to generate long reads then your target may be different from the one you'd select for short reads. Again, the literature is the fastest resource.

Another option is to download VAMPS datasets and mine them: you can see which regions distinguish among which species, the lengths they cover, etc... That is more laborious and difficult.

And in case of doubt, the worst choice, but one that often works reasonably well, is to simply repeat blindly what others have done on similar experiments without trying to understand why they did. You risk making huge mistakes and that a careful reviewer may turn your work down, but you also get chances no one will notice your carelessness.

capsicum · 07-02-2012, 05:26 PM

Quote:

Originally Posted by cliffbeall

One thing you might want to think about is doing more than one region. We looked at V1-2 and V4 on oral bacteria (pre-titanium) and found that V4 significantly undercounted the Bacteroidetes phylum, despite that the primers match Bacteroidetes perfectly well.

The Human Microbiome Project followed a similar tack, with their V1-3 and V3-5 primers. Our current strategy is to use their sequences, on the thought they will become a standard (at least for a few months until the technology changes again).

I think they're using V4 now for Illumina sequencing.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Barcoding 16S amplicons for Illumina Seq	alfredman1	General	8	10-22-2013 05:37 AM
best primers for 16S amplicon sequencing	dca	454 Pyrosequencing	9	09-11-2012 10:49 AM
454 Titanium 16S HMP primers?	cliffbeall	454 Pyrosequencing	9	05-11-2012 06:52 AM
16S metagenome sequencing directions(F/R)	jiaco	454 Pyrosequencing	12	01-26-2012 02:39 PM
Normalization of 16S PCR products	cmazzoni	Metagenomics	0	07-21-2011 06:17 AM

12-07-2011, 07:28 AM	#1
dph Junior Member Location: on earth Join Date: Dec 2011 Posts: 2	best 16s region to amplify with NGS ? Does someone have an idea of the best 16s gene region to amplify with NGS in order to assess microbial (bacterial) diversity in environmental samples ? cheers

04-20-2012, 04:32 AM	#6
jrvalverde Junior Member Location: Madrid, Spain Join Date: Dec 2009 Posts: 4	Rtfp Read the papers. As already pointed out, it depends on what you want. Certain regions work better than others to distinguish among different species, but cannot resolve among others. The best way to know is to find out studies about your target community and see whether they did work. For instance, v6 will not distinguish E coli from Salmonella, which is serious if you want to analyze health data, but may be irrelevant for soil data. A second factor is the technology: if you are going to generate long reads then your target may be different from the one you'd select for short reads. Again, the literature is the fastest resource. Another option is to download VAMPS datasets and mine them: you can see which regions distinguish among which species, the lengths they cover, etc... That is more laborious and difficult. And in case of doubt, the worst choice, but one that often works reasonably well, is to simply repeat blindly what others have done on similar experiments without trying to understand why they did. You risk making huge mistakes and that a careful reviewer may turn your work down, but you also get chances no one will notice your carelessness.

12-08-2011, 10:44 AM	#2
dph Junior Member Location: on earth Join Date: Dec 2011 Posts: 2	no reply ??? no idea ?

12-09-2011, 06:03 AM	#3
themerlin Member Location: Flagstaff, AZ Join Date: Feb 2010 Posts: 51	There has actually been a fair bit of work done on this in the literature. I'd look at some papers out of Rob Knight's group.

12-14-2011, 07:29 AM	#4
cliffbeall Senior Member Location: Ohio Join Date: Jan 2010 Posts: 144	One thing you might want to think about is doing more than one region. We looked at V1-2 and V4 on oral bacteria (pre-titanium) and found that V4 significantly undercounted the Bacteroidetes phylum, despite that the primers match Bacteroidetes perfectly well. The Human Microbiome Project followed a similar tack, with their V1-3 and V3-5 primers. Our current strategy is to use their sequences, on the thought they will become a standard (at least for a few months until the technology changes again).

12-19-2011, 09:58 AM	#5
Jean Member Location: Canada Join Date: Nov 2008 Posts: 37	What environment? What are you trying to find or distinguish? What platform will you be sequencing with? The "best" region is the one that will work for the questions you want to answer. If, for example, you are trying to distinguish two species then one variable region might have more distinguishing power than another.