MOTIVATION: Whole-genome sequencing has revolutionized biosciences by providing tools for constructing complete DNA sequences of individuals. With entire genomes at hand, scientists can pinpoint DNA fragments responsible for oncogenesis and predict patient responses to cancer treatments. Machine learning plays a paramount role in this process. However, the sheer volume of whole-genome data makes it difficult to encode the characteristics of genomic variants as features for learning algorithms. RESULTS: In this article, we propose three feature extraction methods that facilitate classifier learning from sets of genomic variants. The core contributions of this work include: (i) strategies for determining features using variant length binning, clustering and density estimation; (ii) a programing library for automating distribution-based feature extraction in machine learning pipelines. The proposed methods have been validated on five real-world datasets using four different classification algorithms and a clustering approach. Experiments on genomes of 219 ovarian, 61 lung and 929 breast cancer patients show that the proposed approaches automatically identify genomic biomarkers associated with cancer subtypes and clinical response to oncological treatment. Finally, we show that the extracted features can be used alongside unsupervised learning methods to analyze genomic samples. AVAILABILITY AND IMPLEMENTATION: The source code of the presented algorithms and reproducible experimental scripts are available on Github at https://github.com/MNMdiagnostics/dbfe. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: The search for causative genetic variants in rare diseases of presumed monogenic inheritance has been boosted by the implementation of whole exome (WES) and whole genome (WGS) sequencing. In many cases, WGS seems to be superior to WES, but the analysis and visualization of the vast amounts of data is demanding. RESULTS: To aid this challenge, we have developed a new tool-RareVariantVis-for analysis of genome sequence data (including non-coding regions) for both germ line and somatic variants. It visualizes variants along their respective chromosomes, providing information about exact chromosomal position, zygosity and frequency, with point-and-click information regarding dbSNP IDs, gene association and variant inheritance. Rare variants as well as de novo variants can be flagged in different colors. We show the performance of the RareVariantVis tool in the Genome in a Bottle WGS data set. AVAILABILITY AND IMPLEMENTATION: https://www.bioconductor.org/packages/3.3/bioc/html/RareVariantVis.html CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

In scientific data analysis, workflows are used to integrate and coordinate resources such as databases and tools. Workflows are normally executed by an orchestrator that invokes component services and mediates data transport between them. Scientific data are frequently large, and brokering large data increases the load on the orchestrator and reduces workflow performance. To remedy this problem, we demonstrate how plain SOAP web services can be tailored to support direct service-to-service data transport, thus allowing the orchestrator to delegate the data-flow. We formally define a data-flow delegation message, develop an XML schema for it, and analyze performance improvement of data-flow delegation empirically in comparison with the regular orchestration using an example bioinformatics workflow.

Scientific workflow systems facilitate scientific experiments by integrating and coordinating geographically distributed data and algorithmic services in a loosely coupled manner. Most scientific workflow-engines use centralized coordination as the choice of approach for executing workflows, requiring the coordinator (i.e., workflow-engine) to send and receive all input and output data of component services. Such indirect data communication between the component services increases the data-traffic of the coordinator and weakens the performance of the workflow. To optimize this, we propose an approach where data-flow is dynamically delegated from the coordinator to the component services, with direct transportation of data between the component services.