Comparison of whole transcriptome and targeted RNA sequencing for ecological high-throughput transcriptomics

In 2019, in anticipation of a new program of research in ecological high-throughput transcriptomics, US EPA organized a federal government challenge aimed at identifying and evaluating low cost, high-throughput, RNA sequencing technologies that could support the aims of the research. Through a competition open to individuals and non-governmental organizations worldwide, innovators were invited to demonstrate their solutions. Each responding solver was provided a set of nine pooled RNA samples from each of four species of aquatic organisms (n=36 samples total). Three independent Solver teams submitted data, and technology descriptions, for five different analysis methods. Each was evaluated according to a pre-defined scoring rubric that considered accuracy, precision, transcriptome coverage for each species, and commercial viability (cost per sample and throughput considerations). Among the five Solutions submitted, a targeted approach (TempO-Seq) that employed sentinel gene sets representing 5-11% of the whole transcriptome was ranked as the top solution. However, all submitted solutions were determined to be viable for the research and had specific strengths and weaknesses. In a follow up investigation, transcriptomic points of departure based on a sentinel gene set were generally found to fall within a factor of 10 or less of those based on whole transcriptome sequencing. Results support the conclusion that a wide range of sequencing technologies and approaches are suitable for the work. Detailed and transparent reporting of the approaches used will help support uptake in science-based decision-making. The dataset contains the Solver-submitted count matrices; R code for analyses that supported judging; In silico subsets.