Some further comments below, I still need to make suggestions on the statement of need
My read-through notes:
[ ] Affiliations: Doesn't Sofia have 3?
[x] Summary: isn't necessarily correct: can be eukaruotes (not micrbes),
[x] Summary: first sentence is history that doesn't necessarily prime the main question
[x] Summary: Missing wrap up sentence
[x] SoN: Wrong citation MEGAN/MALT
[x] SoN: maybe shortly describe example problem of varying profiler output formats
[x] SoN: Condense other tools sentences into one pargraph, mention kraken2tools(?), QIIME2 is mostly aimed at 16S etc, doesn't CAMI challenge also have a sort of format (more generic)?
[ ] SoN: Missing caveats (which it does/doesn't do, e.g. doesn't do analysis, fixed output to read counts)
[x] Acknolwegments: Max Planck Society
[x] Acknowedgments: RAPID-Seq should have quotes
Questions:
[ ] TAXPASTA or taxpasta or Taxpasta? Should pick one and use everywhere (e.g. docs has TAXPASTA, taxpasta, Taxpasta)
[ ] Do we care/want a simple image summarising the tool, like a graphical summary/figure?
Shower thoughts:
[x] should explicitly mention normally researchers would have to write their own parsers for each and every of the tools
[x] even biom format would require data parting to get it into the right format, AFAICR it didn't have parses for different tools
[x] Krakentools only works for a specific 'family' of profilers (kraken and derivatives)