add the average identity percentage to the counts

[eparejatobes]

From the BLAST output we have the pident field. We just need to

1. add this value to the assignment output (this is useful in itself)
2. calculate the average in the counting step

[laughedelic]

@eparejatobes one thing I'm doubting here:

• LCA can be a node that is not among the BLAST hits, so which pident should correspond to it in the assignment results? average among all hits for this read?

[laughedelic]

Never mind. I forgot that we keep only hits with the same maximum pident 😓

[eparejatobes]

In general it should be average across all hits; we're keeping the max so no need to calculate it, but that could be a future setting: something like calculate the LCA for all hits close to the maximum pident, particularly when setting the pident threshold to a lower value.

[rtobes]

I agree. And when MG7 evolve for other uses (functional assignment for example) it would be useful. (the parameter is pident :))

[laughedelic]

This is done. With average pident:

• in the LCA assignment table per read across all hits
• in the counts table per tax ID