Avoid hardcoding microbe column start point: support arbitrary amounts of sample metadata about qurro

[fedarko]

Yeah this is important now, since it's messing with arbitrary data support. Figure out how many sample metadata columns there are, and then store this in the JSON (or heck even just write it into the JS code directly).

This seems to be causing bugs in searches (e.g. searching for "Bacteria" in the example pseudomonas dataset doesn't hit everything, even though i'm pretty sure all the taxa listed are Bacteria), since some taxa are getting missed.

Also, related?, the line in filterTaxa() that retrieves a list of taxa via Object.keys() seems to be getting 3 extra taxa. See if that's fixed by addressing the above bug, but if not we need to examine that also.

[fedarko]

this has been fixed by #39 being closed!