Capture "pref_name" from ChEMBL

As of v34

```
Pref_name curation. Progress has been made towards standardising drug and clinical candidate pref_names (in MOLECULE_DICTIONARY.PREF_NAME) whereby an approved drug name (FDA /EMA) is assigned in the first instance, if available. If not available, the USAN is assigned, followed by the INN name, respectively. If the USAN/INN name assignment is ambiguous, the FDA GSRS preferred name is used. A company research code, or Clinical Trial intervention name, is assigned if no standardised name is available. For virtual parent compounds, progress has been made towards assigning a distinct pref_name (typically based on the FDA GSRS preferred name) that differs from the child compound name.
```

The issue provides ChEMBL v34 curation guidance but names no repository file, test, or entry point. Start by locating the existing ChEMBL ingestion or molecule-dictionary handling, then define how PREF_NAME should be captured and what validation demonstrates completion.

Written by the indexing model from the issue text.