Odd values for atom pKa

Hi all, thanks for this great package.
I'm getting odd results from the pKa prediction methods `getBasicPkaValue` and `getAcidPkaValue`. It's not clear what values they actually are, but I'll assume the obvious (it returns a pKa for an atom).

The below demonstrates somewhat normal behaviour:
```python
from indigo import Indigo
smiles = 'NCc1ccc(CO)cc1'
ind = Indigo()
ind.setOption("pKa-model", "advanced")
ind.setOption("pKa-model-level", 1)
imol = ind.loadMolecule(smiles)

for atom in imol.iterateAtoms():
print(atom.atomicNumber(), imol.getBasicPkaValue(atom, 5, 0))

```
```
7 9.930000305175781
6 -100.0
6 -100.0
6 -100.0
6 -100.0
6 -100.0
6 -100.0
8 -100.0
6 -100.0
6 -100.0
```
The nitrogen has a pKa around 10 which seems close enough.

however the acid pKa's seem off:
```python
for atom in imol.iterateAtoms():
print(atom.atomicNumber(), imol.getAcidPkaValue(atom, 5, 0))
```
```
7 8.9399995803833
6 6.929999828338623
6 100.0
6 100.0
6 100.0
6 100.0
6 6.929999828338623
8 13.5600004196167
6 100.0
6 100.0
```
setting the minimum level to `2` instead of `0` (as I saw in some examples) leads to all values being 100 or -100.

This leads to a few questions:
- What's the method being used to predict pKas here?
- Does 100 or -100 just mean 'out of range'?
- What is the purpose of the `level` and `min_level` arguments, and why would increasing the `min_level` lead to a broken output?
- Why would some carbons get an acidic pKa around 7?

Thanks for your time