From months to minutes: Open-source web tool moves the needle towards instant microbiome meta-analyses
La Jolla, Calif., October 01, 2018 — Multiomics, the combination of methods that generate data about “omes,” such as the genome, proteome, microbiome, etc. is an emerging approach to microbiome science providing insights into the composition and function of microbial communities one study at a time. In order for scientists to be able to translate findings across populations, they need to be able to see all of the data in one place (referred to as a meta-analysis). Now, researchers at the UC San Diego Center for Microbiome Innovation have published an open-source web tool that enables meta-analyses in minutes—something that would have typically taken researchers months.
|Qiita figure example|
The tool is named Qiita (phonetically pronounced “cheetah”), alluding to its ability to dramatically accelerate the analysis and comparison of microbiome studies.
“Recently, we’ve seen exponential growth in the number of studies that generate large quantities of microbiome and metabolome data, enabled by advances in high-throughput techniques,” said Antonio González, lead developer of Qiita and a bioinformatics analyst in UC San Diego's Knight Lab. “To meet the demand, bioinformatics tools have advanced to enable us to put these samples in the context of other studies.”
Comparing studies gives scientists the big picture, enabling important insights into human health and disease. “In principle, that vast increase in available data should enable broader and more accurate insights into the diversity and functional impacts of the microbial world,” said González. “But unfortunately, these tools require increased time and effort by highly-trained individuals, and we generate data faster than these few skilled individuals can process it.”
Despite the challenges, meta-analyses of microbiome studies have a rich history of success in identifying the drivers of diversity in microbial communities, characterizing the evolution of the vertebrate gut microbiome, and surveying built environments.
Meta-analyses also enable scientists to identify discrepancies in scientific studies, such as different methods of DNA extraction. “We developed Qiita to enable faster meta-analyses that anyone can do,” said González. “Because it’s an open-source web-based platform, non-bioinformaticians can perform their own analyses and meta-analyses in a simple graphical user interface.”
Users start by creating a study within Qiita that contains a description of the work and fill out a template with detailed information about how the samples were collected and processed. Then, they upload the raw data generated by common sequencing platforms. Users can then select how they want the data to be processed, and compare it to hundreds of thousands of samples already in the database.
To demonstrate the utility of the platform, the authors tested the reproducibility of a study of how the microbiomes of people with Inflammatory Bowel Disease (IBD) relate to those of healthy individuals.
“We combined the sequencing data from six studies that contained data from individuals with IBD, patients with recurrent C. Difficile infections that underwent a fecal microbiota transplant and the Human Microbiome Project”, said González. “In less than f
ive minutes, we were able to reproduce the results of the initial studies and create a graph of “healthy” versus “non-healthy” to show what the microbiome in different disease states looks like.”
Even though Qiita was not officially released until this week, it already contains information of more samples than any other publicly-accessible resource—over 50 TB of omics data from over 460,000 samples originating from studies that span the whole world—and has more than 100 citations on Google Scholar.
“We expect that this tool will be widely adopted by microbiome researchers,” said Rob Knight, who directs the Center for Microbiome Innovation. “In fact, that is already the case; we regularly host live training workshops at UC San Diego.”
This isn’t the first time The Knight Lab has developed a tool that has become the standard in microbiome studies as his lab developed Qiime2, a bioinformatics processing pipeline that reached 10,000 citations in the scientific literature last month—a number almost unheard of for a methods paper.
In terms of what’s next, González says the dream is to be able to collect and analyze a fecal sample while one is on the toilet. “Eventually, we’ll be able to tell people whether they’re microbiome is moving towards, or away from, a healthy state on a daily basis.”
The full title of the paper is “Qiita’s web-enabled platform accelerates microbiome meta-analyses from months to minutes” and can be found in Nature Methods.
Additional co-authors are Jose A. Navas-Molina, Tomasz Kosciolek, Daniel McDonald,
Yoshiki Vázquez-Baeza, Gail Ackermann, Jeff DeReus, Stefan Janssen, Austin D. Swafford, Stephanie B. Orchanian, Jon G. Sanders, Joshua Shorenstein, Hannes Holste, Semar Petrus, Adam Robbins-Pianka, Colin J. Brislawn, Mingxun Wang, Jai Ram Rideout, Evan Bolyen, Matthew Dillon, J Gregory Caporaso and Pieter C. Dorrestein.
The work was supported in part by Alfred P. Sloan Foundation 2017-9838 & 2015-13933, NIH/NIDDK P01DK078669, NSF DBI-1565057 & 1565100, Office of Naval Research (ONR) N00014-15-1-2809, and U.S. ARMY CDMRP W81XWH-15-1-0653.
Center for Microbiome Innovation