Representations of molecules can mostly be summed up into two categories: (1) linear notations, which use chemical rules to create string representations of the molecule, and (2) graphs, which represent atoms as nodes and bonds as edges. Of the two, linear notations are more compact, require relatively little disk space to store, and tend to scale linearly as the molecule size grows. However, graph representations offer more versatility and can potentially capture more features. There are multiple types of linear notation generation, with the SMILES method being the most popular due to its simplicity. In brief, the notation is created by assigning a number to each atom in the molecule and traversing the molecule like a graph using depth-first-search.