Parameterized pattern queries

We introduce parameterized pattern queries as a new paradigm to extend traditional pattern expressions over sequence databases. A parameterized pattern is essentially a string made of constant symbols or variables where variables can be matched against any symbol of the input string. Parameterized patterns allow a concise and expressive definition of regular expressions that would be very complex to describe without variables. They can also be used to express additional constraints, to “relax” pattern expressions by allowing more freedom, and finally to cluster patterns in order to minimize the number of symbols’ comparisons.