this is alpha-quality software
This package extends the counting renaissance methods of Lemey et. al. 2012 to estimate site-specific selection from pairwise alignments of deep sequencing data under varying coverage. An application to B cell receptor sequences is described in our preprint on arXiv.
Input sequences should be in BAM format. A site-specific HKY JSON model specification is required - these can be generated via fit-star.
Sequences are processed in parallel using Apache Spark.
BEAGLE is required for likelihood calculations.
See the deploy/ directory for an example of provisioning a cluster on EC2.
startreerenaissance is a mix of Java and Scala; built with sbt.
You may build a .jar file with application code and dependencies using sbt assembly.
Unit tests are run via sbt test.
Code in startreerenaissance is licensed under the GNU Public License.
Licenses of dependencies are included in lib/.