Bio4j has an intermediate Blueprints layer, which allows us to make a default implementation of the abstract domain model using Tinkerpop Blueprints API and at the same time stay independent from the choice of database technology.
See also technology specific implementations:
To use it in your sbt-project, add this to build.sbt:
resolvers += "Era7 maven releases" at "http://releases.era7.com.s3.amazonaws.com"
libraryDependencies += "bio4j" % "blueprints" % "0.4.0"