How many reads/sample is required for species level identification when performing 16S rRNA sequencing of the gut microbiome?
Dear Stefanie,
We used in our study 16 individual birds per treatment in a study and detected some differences in microbiome. We and also other authors obsersved very high variation between birds, so it is important not to pool samples of birds!
Kind regards,
Evelien
Thanks Evelien. What techniques did you do to investigate the microbial composition, next generation sequencing? If so, how many reads per sample did you generate and were you able to compare species levels or only Family/Genus?
Dear Stefanie,
we used a chicken intestinal microarray, with oligonucleotides based on the 16S rRNA gene sequences and Enterobacteriaceae hsp60 gene sequences. See for details the attached paper.
Oh ok. Yes my question relates to 16S rRNA sequencing and read numbers so unfortunately it will not help me.
Hi Stephanie,
this depends on the expected diversity. If you expect a complex microbiome you will need more reads to get a high coverage of your samples than if you are investigating a microbiome with few species. If you expect few species in the community (what might be the case if you are investigating the gut community of f.e. small insects ), a lower number of total reads will be sufficient. You can calculate rarefaction curves and diversity estimators after sequencing to decide if seq depth was enough or if you want to resequence samples to improve the coverage.
Beside of that, amplicon sequencing with 150-500bp read length can reliably predict genus level but can give you just an estimate for species affiliation (f.e. by blasting against 16S rRNA gene databases). This is not dependent on the number of reads you get, as amplicon sequencing is based on the amplification of a certain region. Increasing the number of reads can increase seq coverage but not taxonomic resolution (species, genus, family...)
Hope this helps,
best, Evelyne
In addition to what Evelyn and Nir said...
In my opinion, you should know the objectives of your study... If you want to investigate and determine species level identification of microbiota using NGS it is not possible because Roche 454 only capture about 600 bp and Miseq can capture about 200 bp.. As we all know to get a full length 1500 bp makes a reliable species identification. Unless you use the PacBio chemistry you can capture 1500 bp and with greater amount of reads per sample too. I suggest you can use Miseq so that the rarefaction curve will achieve a plateau then perform upto genus level classification.. If you find any genus/genera that is noteworthy then perform additinal experiment using full length16s rRNA gene probes. Or you can do other things as well... In any case, you should find a good methodology to answer your research questions.
Hope my answer help... Happy researching!
Dear Stefanie Malan,
I concur with Evelyne Mann, approx 1.5 kbp size of 16S through Sanger sequencing can provide you with the question you are seeking. However, due to limitation in current next gen sequencing technology (considering Illumina is the choice world over) we had to limit our-self with one or combination of variable regions within 16S (V1-V9) which the sequencing chemistry allows.. Typically V4 or V3-V4 is the choice (you can find publication which ones were the chosen one). As such, we have to moved from the species concept to the OTU concept (Pat Schloss explains the "phylotype concept" in MOTHUR documentation if you are interested). As rightly put by other researchers, inferring species level taxonomy will be questionable by just sequencing a small part of the 16S. So, tagged amplicon sequencing is great for community level understanding.
Dear all,
I'm working with humans and I got 10,000 reads at specie level. I'm using the Miseq plataform and we sequenced the the V3-V4 region. Can I say that I have a specific bacteria with these number of lectures?
- Badges
- Science topic