LAUSR

Omics research inevitably involves the collection and analysis of big data, which can only be handled by automated approaches. Here we point out that analysis of big data in the field of genomics dictates certain requirements, such as specialized software, quality control of input data, and simplification for visualization of the results. The latter results in loss of information, as is exemplified for phylogenetic trees. Clear communication of big data analyses can be enhanced by novel visualization strategies. The interpretation of findings is sometimes hampered when dedicated analytical tools are not fully understood by microbiologists, while the researchers performing these analyses may not have a full overview of the biology of the microbes under study. These issues are illustrated here, using SARS-Cov-2 and Salmonella enterica as zoonotic examples. Whereas in scientific communications jargon should be avoided or explained, nomenclature to group similar organisms and distinguish these from more distant relatives is not only essential, but also influences the interpretation of results. Unfortunately, changes in taxonomically accepted names are now so frequent that they hamper rather than assist research, as is illustrated with difficulties of microbiome studies. Nomenclature to group viral isolates, as is done for SARS-Cov2, is also not without difficulties. Some weaknesses in current omics research stem from poor quality of data or biased databases, and problems can be magnified by machine learning approaches. Moreover, the overall opus of scientific publications can now be considered 'big data', as is illustrated by the avalanche of Covid-19-related publications. The peer-review model of scientific publishing is only barely coping with this novel situation, resulting in retractions and publication of bogus works. The avalanche of scientific publications that originated from the current pandemic can obstruct literature searches and this will unfortunately continue over time.