The Joint Genome Institute plans to transition from a sequencing center to a "genomic foundry" — a one-stop shop for large-scale functional annotation, single-cell genomics and transcriptomics, high-throughput custom sample prep, and analysis expertise, among other proposed services, said Eddy Rubin at JGI's sixth annual User Meeting in Walnut Creek, Calif. Ultimately, Rubin said, he and his colleagues aim to "transform the JGI from largely a sequence production facility to the 'DOE Genomic Foundry,' pioneering the use of functional genomics capabilities to solve energy and environmental problems." In the future, rather than sending samples into JGI, Rubin envisions researchers approaching the institute with project proposals; JGI, Rubin said, exists "to enable the science of users — to push technologies and help users to do science in advance ways. … We need to figure out what the JGI needs to be in five years."
While not set in stone, Rubin presented potential "stretch goals" for JGI. By 2016, he said, the institute intends to "design and build a genome to address a whole-genome hypothesis." Through 2021, JGI would like to "routinely assign a function to [more than] 90 percent of the genes in a microbial [or] plant genome," ascertain "new branches of life in the dark matter of unculturable organisms," and "characterize and model complex environmental systems to a level where we can correctly predict response to changes," Rubin said.
"Sequencing is meant for a centralized facility," he added. But as individual sequencing projects become more specialized — as they "branch father away from the trunk" — the functional analyses they need often do as well. "The point of the sequencing is to generate lots of sequence; the point of the [functional analysis] is to generate functions for thousands of genes," Rubin said. To that end, JGI would need to come up with an approach "that puts it [functional analysis] back into a fully high-throughput function and takes it out of the sort of 'boutique model,'" he said. "In reality, biology hasn't come up with many things that sort of need centers. Biology is an R01 process — generally — and genome centers were the one place that ... strength came in numbers. ... Early on in genomics they didn't even think that we could scale," he said, adding that perhaps the same could one day be true for functional annotation.
Rubin said the JGI welcomes additional user feedback as the institute works to refine its strategic plan.