Statement of Work for Diet Reconstruction
August 29, 2014
We will use a multiplexed PCR technique to reconstruct diets from the analysis of DNA in fecal samples. DNA will be extracted from all samples and PCR-amplified using primers that contain error-correcting barcodes and Illumina adapters. After amplification, we will confirm amplicon lengths by gel electrophoresis, quantify amplicon concentrations, purify the amplicons, and normalize the amplicons for sequencing on an Illumina MiSeq instrument. Sequence data are run through a bioinformatics pipeline that classifies sequences against a taxonomic database to which we have added any plant reference samples. Results are provided as a list of all plant species found in each sample and the relative abundances of those species (based on sequence counts).
Plants in diet are limited to higher plants. For plants, we rely on a portion of the chloroplast trnL intron using the c and h trnL primers. The CO1 sequence is targeted to distinguish insects in diets. Due to differences in lengths of the trnL and CO1 sequences, trnL and CO1 sequences will be sequenced separately from one another. ITS primers are used for fungi.
Closely-related plant or animal species may not be able to be distinguished. This uncertainty will be noted in the data table.
Raw sequence data and reference sequences for any reference samples from plants supplied by the customer will be made available for customers.
Due to inherent difficulty in isolating and amplifying DNA from environmental samples, not all samples will yield sufficient amounts of DNA for sequencing. Even for high quality samples from which we can extract DNA, we typically fail to get sufficient data from at 10% of samples and this failure rate goes up appreciably for samples that are difficult to amplify or otherwise problematic. Any samples for which DNA amounts are insufficient for sequencing, the customer will only be charged a partial rate.
Depending on availability of reagents, quality of extracted DNA, and availability of the sequencing instruments, analyses can take 1-3 months to complete upon receipt of the samples. Many of these factors are beyond our control but we will work to process the samples as soon as possible.
A minimum of 1000 sequences are estimated to be generated per fecal sample, but this varies on the quality of the sample.