To conduct a soil experiment on microbial composition as well as mycorrhizal fungal composition, what are the advantages/ disadvantages of using illumina over TRFLP?
TRFLP is still used, but is increasingly viewed as old-fashioned. We have had several manuscripts criticized rather heavily because of the use of TRFLP methods, so have now switched completely to MiSeq amplicon sequencing.
The issue is, obviously, that TRFLP gives you only the 40 or so dominant phylotypes (but no phylotyopic information), so while its simple, cheap and rather good for comparative analyses, you are ignoring most of the diversity.
You will also be aware that there is a move from amplicon (16S, ITS etc) sequence analysis to full metagenome sequencing, and the extraction of phylotypic data (and much more, of course) from the very large metagenomic sequence datasets. Needs some more computational and bioinformatics capacity, but avoids PCR biases and provides access to a raft of gene diversity, metabolic reconstruction, genome re-assembly etc options.
TRFLP is a good method for looking at patterns at a reduced cost. I have been able to get out 540 TRFLP profiles for less than the cost of 42 samples through pyrosequencing. You are limited to just looking at patterns and have no phylogenetic information with TRFLPs.
Like most here already said TRFLP is cheaper, but less informative than sequencing. Currently methodologies based on target sequencing are used more and much more informative.
I personally think NGS is much better than TRFLP because NGS give you much more both quality and quantity of your full genomic DNA. TRFLP as you know its more focus on Restriction enzyme site and because you're more interested on matagenomics of your soil microbial community, it might happen that more than one micribe has the same RES and on your gel and wont be true reflection of your community. NGS is a bit expensive but I personally prefere it!