In silico restriction analysis for identifying microbial communities in T-RFLP fingerprints

We present here a technique for correlating T-RFLP fingerprints with genomic clone libraries, using a in silico restriction analysis, in order to identify the microbial (Bacteria and Archaea) communities present in an environmental sample. This technique is very useful for independent of culture metagenomic studies. We also show some results on the application of this technique to 3 environmental samples from an anthropogenically extreme site. We confirm that T-RFLP results are quite reproducible both in peak’s location and area. We found that the peak position (Terminal Restriction Fragment, TRF) can be identified with an uncertainty of only 0.3 base pair, while the percentage o fluorescence (frequency of population) has about 4% of relative uncertainty. Using the TRFs obtained from the in silico restriction as reference, we found that deletion and insertion during electrophoresis step of the T-RFLP and cloning must be taken into account. They typically produce a shift in the range [−2 to +1] in PCR/cloning/sequencing and [−4 to +1] in PCR/T-RFLP, independent of fragment size. This also means that deletion is more usual than insertion. The in silico restriction analysis allowed us to recognize 100% of the T-RFLP peaks of abundant populations, 60% of intermediate and 50% of poorly represented ones. Also, almost all populations recovered in the clone library could be associated to T-RFLP peaks, but not vice-versa , confirming that the T-RFLP is an efficient technique for detecting the less dominant populations of a microbial community.