I read some criticisms on the limitations of the Illumina protocol due to the use of a single primer. I am new in the use of the technique thus any suggestions are very welcome.
Hello Ana, multiple primers can be used in Illumina sequencing, though people prefer V4 region now. I wonder which articles have you read. Nowadays, many sequencing companies can provide multiplex PCR-Illumina sequencing service, offering information of V3/V4/V6 region of 16S rRNA gene.
Besides 16S rRNA gene, I'd like to suggest GeoChip for your study.
Metagenomics is the study of genes and genomes from multiple populations. The approach to identify the microbes that are present and at what abundance is typically referred to as phylogenetic marker gene sequencing.
Either way you will need to extract nucleic acids. This can be done using a commercially available kit (e.g. Powersoil from Mobio). The next step depends whether you want to do metagenomics (shotgun sequencing) or phylogenetic marker gene sequencing (amplicon sequencing).
Obviously if you opt for amplicon sequencing the breadth of populations that you will be able to detect is dependent on the primers that you use. For bacteria and archaea you can amplify 926-1492 and then use the MiSeq platform for sequencing. You can run more than one sample on the same run by adding a barcode to each primer. You can order your own barcoded primers or ask a sequencing company to add this prior to pooling. You can then use the QIIME pipeline to process your sequence data. Be warned that processing these datasets is not trivial and typically requires access to supercomputing infrastructure. For eukarya you can look into universalish primers that target 18S rRNA but this will only enable identification to phylum/class level. Alternatively you can target other regions such as ITS (e.g. for fungi).
Shotgun metagenomic sequencing may facilitate characterization of a wide-range of different groups but your ability to detect rarer populations will be compromised relative to amplicon sequencing unless you massively increase the sequencing depth of each sample. You will need at the very minimum 1/3 - 1/2 of a lane of Illumina HiSeq per sample to make much sense of your data. The level of bioinformatics require to analyse metagenomic datasets is typically advanced and definately requires serious computing power.
If you could be more specific about the groups that you would like to characterise then you will almost certainly receive more definitive feedback such as protocols.
We usually recommend to our customers, 16S sequencing (bacteria/archea) and ITS (fungi) sequencing using the 454 or MiSeq platforms. In the case of 454 we use V1-V3 and for MiSeq we follow Illumina's protocol using V3-V4 PE so you end up with a 550bp which is good for identification. Ion Torrent is not really recommended. QIIME is good but if you have no programming background it is complicated and time consuming.
But for the analysis you should always consider what database you use for identification. A good recommended database is EzTaxon, which includes cultured and uncultured species, you can find more info on wikipedia.
If microbial diversity is your unique interest, I would recomend the analysis of 16S rRNA using different primers covering some variable regions. I do not think the Illumina technique has limitations. The 16s analysis per si has some limitations for classification but it is ok in the study of overall diversity in the genus level.
I had done some seq. part for V3-V6 region in 16S rRNA. The system what we are using was Ion torrent. the amplicon size was around 100 bp(sogin at al). we worked on both water and soil samples. time taking part is analysis which i recommended u MG-RAST online software. you can find easily on internet.
Thank you very much for your most helpful answers.
My goal is to identify the biodiversity of soil microbiology.
The only main target I have are nitrogen fixing bacteria, however I would not like to be restricted to those, instead I would like to be able to identify the broadest bacterial range possible in my samples.