Table 1 Issues and potential solutions to data cohesiveness in respiratory microbiota studies

Data accessibility

Ethical challenges

Consider deposition of data in original research proposals

Develop and support the use of guarded archives with established regulatory oversight. Develop patient communication plans that could be used when consenting individuals in order to inform permissions for future data use

What to upload?

Raw sequence data

Clearly labelled controls

Minimal metadata^a

Cohort heterogeneity

Diagnostics

Clearly state diagnostic criteria for cases

Where possible, report symptoms or tests used for decision-making

Determine how accurately current diagnostic criteria reflect pathobiological processes and determine if any can be combined from a microbiota perspective

Demographics

Report key demographic variables^b

Try to recruit diverse cohorts

Further research is needed to determine key demographic variables that may confound meta-analyses

Antibiotics

Create a standardised (and granular) reporting metric for reporting recent antibiotic exposure

Improve understanding of respiratory microbiota recovery from antibiotics

Sample collection

Anatomical site

Address knowledge gaps in lower airway sample literature and (where possible) collect paired upper- and lower-airway samples

Improve understanding as to how related the microbiotas of different respiratory anatomical sites are in different conditions or clinical contexts (e.g. during long-term antibiotic therapy)

Sampling method

Clearly state sample method used

Determine consistency of different sampling approaches used for the same anatomical site (e.g. nasal swabs and nasal aspirates; sputum and induced sputum; BAL vs protected brushing)

Sample contamination

Include and report results of negative extraction controls and methods used to mitigate the presence of kitome among low bacterial density samples

Sample processing

DNA extraction

Laboratory protocols should be reported consistent with STORMS and STROBE-metagenomics requirements [11, 12]

Determine the effects of different extraction methods on perceived microbiota profiles for different anatomical sites

PCR bias in amplicon sequencing studies

Determine the effects of different sequence regions on perceived microbiota profiles

Bioinformatics processing

Clearly state software, version numbers, and where possible include analysis scripts as supplementary files when submitting primary data for publication

Make data available for secondary analyses which may use a new pipeline