*3.1. Drug-Likeness of* D. orbita *Secondary Metabolites*

Using the online chemoinformatics software Molinspiration (version 2011.06) the drug-likeness (Table 1) and bioactivity scores (Table 2) are predicted for the main secondary metabolites from *D. orbita*. Drug-likeness is based on Lipinskis "Rule of 5" [69], which considers whether various molecular properties and structure features of a particular molecule are similar to known drugs. These properties, such as hydrophobicity, electronic distribution, hydrogen bonding characteristics, molecule size and flexibility (Table 2), influence the bioavailability, transport properties, affinity to proteins, reactivity, toxicity, metabolic stability of the molecule and thus potential for use as a pharmaceutical drug. Of all the indole derivatives examined (Table 1a), only a single violation of the rule of 5 was found. This was for tyriverdin, due to a molecule weight exceeding 500 mass units (Table 1a). As expected, choline and all of the choline esters conform to the rule of 5 for drug-likeness (Table 1b).



(**B**) # Larger value bioactivity scores indicate a higher probability that the molecule will be active; \* potential activity; \*\* high potential activity.

The drug-likeness for Tyrian purple maybe over-estimated. The log *p* values for 6,6-dibromoindigo is perhaps lower than anticipated considering the fact that this compound is highly non-polar and generally insoluble at room temperature in all organic solvents [2]. Tyrian purple can only be extracted out of tissue or cloth using hot (>100 ◦C) DMF or DMSO. It appears to form dimers or higher polymers due to the van der Waals attraction between bromine atoms [75], which contribute to the high stability of the compound, but nevertheless the low solubility makes it an unlikely drug candidate. Despite the same log *p* value (Table 1a) 6,6 dibromoindirubin appears to be slightly more soluble in non-polar solvents at room temperature (pers. obs.), perhaps due to reduced polymer formation in this isomer.
