Taxonomy

Users might encounter a number of different taxonomic issues while formatting their data for PHAROS. Some issues identified in beta testing are described below:

Identification above the species level

Species-level identification isn't always possible: for example, species concepts are contentious and in flux for many microbial organisms, and many diagnostic methods work above the species level. It's always okay to use higher-level taxonomic identification in PHAROS; for example:

Host species	Host NCBI tax ID	Pathogen	Pathogen NCBI tax ID
Desmodus rotundus	9430	Alphacoronavirus	693996

For novel viruses, including a GenBank accession can help efforts to re-assign these records to species later on, or researchers can simply update their datasets.

Identification to mixed taxonomic levels

Some datasets will contain results with multiple taxonomic levels, even for the same sample. This is easily managed by splitting records across rows.

If this solution is used, researchers should be careful to split count data in a non-overlapping way. For example, if a baboon fecal sample contains 10 nematode eggs, of which 3 can be identified as Strongyloides sp., this should be entered as:

	Sample ID	Animal ID	Host species	Pathogen	Pathogen NCBI tax ID
1	PapioWorms24	Paan1	Papio anubis	Nematoda	693996
2	PapioWorms24	Paan1	Papio anubis	Strongyloides	6247

	Detection outcome	Detection measurement	Detection measurement units
1	Positive	7	egg count
2	Positive	3	egg count

Uncertain identification

Host identification based on morphology (i.e., without molecular confirmation) may be unreliable for some taxa. Researchers can flag this in the "Detection comments" field.

Mismatch between target and organism

In some cases, detection target and pathogen identification may be misaligned. Researchers should aim to capture molecular methods, or - for visual studies - study scope, as accurately as possible.

Animal ID	Host species	Detection target	Detection outcome	Pathogen
Pl001	Peromyscus leucopus	Arthropoda	Positive	Ixodes scapularis
Pl002	Peromyscus leucopus	Arthropoda	Negative

Taxonomic changes

At the moment, PHAROS is not very robust to changing host or pathogen taxonomy - especially rapidly-changing viral and bacterial taxonomy. The NCBI taxonomic backbone uses unique identifiers to track changes over time, and we suggest that if possible, users fill in these fields ("Host species NCBI tax ID", "Detection target NCBI tax ID", and "Pathogen NCBI tax ID") to make their data more robust to these changes. Eventually, we hope to add back-end integration to update names automatically.