“Available upon request”: not good enough for microbiome data!

Abstract

Open data that is free and publicly available without restrictions is critical for progress in any scientific discipline and has been the cornerstone of sound and reproducible genomics research. Microbiome research is still a relatively young, thriving, active research field, with great biomedical potential. As a large data-driven research field, microbiome projects can include hundreds or even thousands of participants, samples, and associated background ("metadata") parameters. Processing this data, identifying meaningful associations, and determining significance depends on complex, often non-standardized bioinformatics and biostatistics protocols. Reproducibility, transparency, and expandability of these protocols to review, evaluate, and build upon this work is crucial to fulfill on the promise of microbiome research and maintain credibility. At the absolute minimum, unrestricted access to the raw sequencing data and associated metadata is needed and has been recognized and implemented by the scientific community, some journals, and funding agencies. In practice, access to open protocols for data processing and analysis is also important to promote reproducibility and advances in the field but rarely provided. Unfortunately, there appears to be an increasing number of studies that are failing to satisfy even basic, community-accepted standards. Motivated by a number of recent negative experiences in our own research projects, as well as our interaction with authors aiming to publish in Microbiome, this editorial aims to shed light on common problems in the field and make recommendations to reinforce a culture of open data and protocols for microbiome research.

Toxicomicrobiomics: The Human Microbiome vs. Pharmaceutical, Dietary, and Environmental Xenobiotics

“…Microbiome studies should include populations with divergent geography and ethnicity (Lloyd-Price et al, 2017). Data and metadata of microbiome studies should be freely accessed over the time, according to peer-reviewed journal guidelines, to allow standardization and validation of bioinformatics and statistical analyses (Langille et al, 2018).…”

Toward unrestricted use of public genomic data

“…Ultimately, the intention of the funding agencies who require pre-publication data sharing has always been to encourage the use of such data by the entire community and to encourage open competition to accelerate discovery and maximize the benefit for members of society who are paying for data generation. And most importantly, as was recently argued, freely available public data without restrictions is considered "critical for progress in any scientific discipline and has been the cornerstone of sound and reproducible genomics research" (11).…”

“…Finally, journal publishers also need to revisit their publication policies, with respect to the availability of the data when a manuscript is submitted for publication. Following the recommendations of Microbiome's editors, the sequence data and their associated metadata need to be freely available together with detailed protocols at the time a manuscript is submitted for peer review, rather than post-publication (11).…”

Towards reproducible, transparent, and systematic benchmarking of omics computational tools

“…While the vast majority of surveyed benchmarking studies are widely disseminated benchmarking data, only 42% of the studies surveyed completely shared benchmarking data (Table 1). Most studies adopted the 'shared upon request' model, which is a less reliable and less reproducible method of data dissemination as it relies on individual authors' availability to perpetually share data 53,54 .…”