Back to All Events

Dipping into Guacamole — a Spark-powered Somatic Variant Caller

  • Phosphorus 1140 Broadway New York, NY, 10001 United States (map)

Next generation sequencing of tumor DNA and RNA has revolutionized cancer genomics, and projects such as the Cancer Genome Atlas have sequenced over ten thousand patient samples. Detecting cancer mutations from paired tumor/normal sequencing is more challenging than traditional germline variant calling because tumors are heterogeneous mixtures of cancer clones.

In this talk, Tim O'Donnell and Ryan Williams from Hammer Lab ( (a lab within the Icahn Institute ( at Mount Sinai ( will walk through development progress on Guacamole (, a somatic variant caller (which helps identify DNA mutations from Next Generation Sequencing data) that combines evidence from multiple DNA or RNA samples from the same patient for better sensitivity.

Guacamole runs on Apache Spark using a novel collection of reusable primitives for distributed operations on aligned short reads. Tim and Ryan will show some examples of these primitives and will also talk a bit about their experience debugging Spark jobs.

We'll have chips and guacamole for eating as well!