Daylight Software in the Avalon Data Warehouse

• Lexical Representations

  • various levels of canonical SMILES great for ‘group by structure’
  • unique constraints on SMILES simplify database loading
  • SMILES and Fingerprints are ASCII text. This makes loading data much easier.

• Structure Search in DayCart

  • high-speed substructure search
  • selective and non-selective queries work well when only first rows are needed
  • multiple substructure queries (optionally with NOT logic) are easy to implement (‘carbonic acids but no diacids’)
  • SMARTS strings are good for saving queries within XML

Notes: