Data from: Next-generation sequencing for rodent barcoding: species identification from fresh, degraded and environmental samples
Data files
Feb 05, 2013 version files 12.48 MB
-
Fasta S1 COI115Species from GenBank.fas
-
Fasta S10 Predator Haplotypes.fas
-
Fasta S2 Cytb115Species from GenBank.fas
-
Fasta S3 Design primers.fasta
-
Fasta S4 Reference Samples.fas
-
Fasta S5 Rodentia Haplotypes.fas
-
Fasta S6 Prey Haplotypes.fas
-
Fasta S7 Museum Murinae True Haplotypes.fas
-
Fasta S8 Museum Human DNA contamination Haplotypes.fas
-
Fasta S9 Museum DNA contamination Haplotypes.fas
-
Fatsa S11 Environnemental contamination Haplotypes.fas
Abstract
Rodentia is the most diverse order among mammals, with more than 2,000 species currently described. Most of the time, species assignation is so difficult based on morphological data solely that identifying rodents at the specific level corresponds to a real challenge. In this study, we compared the applicability of 100 bp mini-barcodes from cytochrome b and cytochrome c oxidase 1 genes to enable rodent species identification. Based on GenBank sequence datasets of 115 rodent species, a 136 bp fragment of cytochrome b was selected as the most discriminatory mini-barcode, and rodent universal primers surrounding this fragment were designed. The efficacy of this new molecular tool was assessed on 946 samples including rodent tissues, feces, museum samples and feces/pellets from predators known to ingest rodents. Utilizing next-generation sequencing technologies able to sequence mixes of DNA, 1,140 amplicons were tagged, multiplexed and sequenced together in one single 454 GS-FLX run. Our method was initially validated on a reference sample set including 265 clearly identified rodent tissues, corresponding to 103 different species. Following validation, 85.6% of 555 rodent samples from Europe, Asia and Africa whose species identity was unknown were able to be identified using the BLASTN program and GenBank reference sequences. In addition, our method proved effective even on degraded rodent DNA samples: 91.8% and 75.9% of samples from feces and museum specimens respectively were correctly identified. Finally, we succeeded in determining the diet of 66.7% of the investigated carnivores from their feces and 81.8% of owls from their pellets. Non-rodent species were also identified, suggesting that our method is sensitive enough to investigate complete predator diets. This study demonstrates how this molecular identification method combined with high-throughput sequencing can open new realms of possibilities in achieving fast, accurate and inexpensive species identification.