Scaling machine learning for target prediction in drug discovery using Apache Spark

Dries Harnie, Mathijs Saey, Alexander Vapirev, Jörg Kurt Wegner, Andrey Gedich, Marvin Steijaert, Hugo Ceulemans, Roel Wuyts, Wolfgang De Meuter

Research output: Contribution to journalArticle

31 Citations (Scopus)


In the context of drug discovery, a key problem is the identification of candidate molecules that affect proteins associated with diseases. Inside Janssen Pharmaceutica, the Chemogenomics project aims to derive new candidates from existing experiments through a set of machine learning predictor programs, written in single-node C++. These programs take a long time to run and are inherently parallel, but do not use multiple nodes. We show how we reimplemented the pipeline using Apache Spark, which enabled us to lift the existing programs to a multi-node cluster without making changes to the predictors. We have benchmarked our Spark pipeline against the original, which shows almost linear speedup up to 8 nodes. In addition, our pipeline generates fewer intermediate files while allowing easier checkpointing and monitoring.
Original languageEnglish
Pages (from-to)409-417
JournalFuture Generation Computer Systems
Publication statusPublished - 24 May 2016


  • Drug discovery
  • Apache Spark
  • Machine learning
  • Target prediction

Fingerprint Dive into the research topics of 'Scaling machine learning for target prediction in drug discovery using Apache Spark'. Together they form a unique fingerprint.

Cite this