[1822] Customized Next Generation Sequencing Tail-End Data Analysis Pipeline for Clinical Cancer Genomics

Somak Roy, Mary Beth Durso, Abigail Wald, Yuri Nikiforov, Marina Nikiforova. University of Pittsburgh, Pittsburgh, PA

Background: Several bioinformatics applications are currently available for processing NGS data. Certain limitations however, preclude the use of such applications in a clinical molecular laboratory. Limited support for 1) customized annotation and platform specific error detection. 2) Customized report generation and ability for cross-talk with existing anatomic pathology laboratory information systems (APLIS) 3) Operating system (OS) commonly used in a clinical laboratory. We attempted to develop a data analysis pipeline to address the above limitations.
Design: TART (Tail-end Analysis and Reporting Tool) is a web-based application developed using ASP.NET version 4.5 (Microsoft, Mountain View, CA). The client-side was scripted using HTML5, CSS3 and JavaScript. The server-side processing relies on interaction with a customized Microsoft SQL server® 2008 R2 (Microsoft, Mountain View, CA) database. This database comprises of three important components – 1) sample and run information 2) Platform specific sequence variation (SV) information and 3) two publically available SV databases; COSMIC (Wellcome Trust Sanger Institute, England) and dbSNP (NCBI). TART is currently designed for processing variant calls generated by targeted next generation sequencing on Ion Torrent PGM.
Results: TART accepts tab-delimited text files, generated by the Variant Caller plugin, as required input. Both variant calls (VC) as well as coverage depth for each hotspot in the mutation panel is processed. Each VC is classified into one of five report levels (1- SV of known clinical significance (CS), 2 – SV of unknown CS, 3 –questionable SV, requiring specific attention, 4 – synonymous SV and 5 – platform specific errors). After pathologist review, the application initiates a “training mode” in response to any change in VC interpretation. As more samples are processed and reviewed by the pathologist, the application “learns” modified or new SV which enriches the database. The report is generated on the server-side in HTML or OpenXML format allowing compatibility with APLIS. The program also allows search and retrieval of archived cases for laboratory quality assurance (QA) activities and research.
Conclusions: This web-based application facilitates rapid and customized VC processing and report generation for targeted next generation sequencing in clinical laboratory. Important advantages include customized annotations, ease of importing reports into APLIS, production of enriched and better annotated SV database, Microsoft Windows OS-support, and report archival and retrieval system.
Category: Special Category - Pan-genomic/Pan-proteomic approaches to Cancer

Tuesday, March 5, 2013 9:30 AM

Poster Session III # 233, Tuesday Morning


Close Window