Database uses too much space

[AlessioMilanese]

Test with building a database with 33k genes of length ~ 800 nucleotides.

To run create_db it takes:

Operation	Time
Load tax	0.3 s
Load alignment	1 m
Check taxonomy	0.3 s
Train all classifiers	12 m
Learn tax level function	30 m
Save file	8 s

When running with time we have:

real      6182.84
user      8619.94
sys        389.02
3167723520  maximum resident set size
         0  average shared memory size
         0  average unshared data size
         0  average unshared stack size
 143395164  page reclaims
        14  page faults
         0  swaps
         0  block input operations
         0  block output operations
         0  messages sent
         0  messages received
         0  signals received
      2279  voluntary context switches
   7431869  involuntary context switches

The file created (HDF5 format) is 400 MB, which is unusable.

When compressed into .zip it's 2.3 MB.

[AlessioMilanese]

We compress the result in ae547b0 (following h5py guide).
The database goes from 402 MB to 40.2 MB.