Graph theoretical approach to structure-activity studies: search for optimal antitumor compounds

Area of Science:

• Computational chemistry
• Medicinal chemistry
• Graph theory

Background:

• Drug discovery involves screening numerous compounds.
• Quantitative structure-activity relationships (QSAR) are crucial for identifying potent drugs.
• Graph theoretical methods offer novel ways to represent and compare molecular structures.

Purpose of the Study:

• To outline a graph theoretical approach for identifying potent drug candidates.
• To quantitatively assess structural similarity among drug molecules.
• To apply the method to a set of antitumor phenyldialkyltriazenes.

Main Methods:

• Identifying a strategic molecular fragment.
• Describing the fragment using graph theoretical invariants, specifically path numbers from weighted bonds.
• Quantitatively deriving molecular similarity based on atomic path numbers.
• Applying the method to known antitumor phenyldialkyltriazenes with reported activity (log(1/C)).

Main Results:

• The graph theoretical approach successfully identified the most potent compound within the tested set.
• The most potent drug identified was 1-(4-NHCOCH3-Phenyl)-3,3-dialkyltriazene.
• The method efficiently narrowed down the search from numerous candidates.

Conclusions:

• Graph theoretical methods provide an effective strategy for drug potency prediction.
• The use of path numbers as invariants allows for quantitative similarity assessment.
• This approach can accelerate the identification of lead compounds in drug discovery.