cdc_titer_column_map: set sr_lot to serum_id

The CDC uses the lot numbers in the `sr_lot` column to discuss titer
measurements, so it makes more sense to use the lot numbers as the
`serum_id`.

This commit also removes the mapping for `sr_ferret` to `ferret_id`
because the `read_flat` method within `tdb/parse` assigns this
`ferret_id` to the `serum_id`, which would overwrite the lot number.
Since we don't need the ferret id as the serum id anymore, there's no
reason to include it in the mapping.

Resolves #126.

source-data/cdc_titer_column_map.tsv

@@ -3,8 +3,7 @@ ag_cdc_id	virus_cdc_id
 ag_strain_name	virus_strain
 ag_passage	virus_strain_passage
 sr_strain_name	serum_strain
-sr_ferret	ferret_id
-sr_lot	lot_number
+sr_lot	serum_id
 sr_passage	serum_antigen_passage
 test_date	assay_date
 test_subtype	subtype