summary - universitat de valènciavarnau/mgvizce.pdf · (bokeh, crossfilter, flask and reportlab)...

MGvizCE: Clinical Exome QC and Analytics Pablo Marin-Garcia, Daniel Perez-Gil, Cristian Perez-Garcia, Alba Sanchis-Juan,Azahara Fuentes, Jose M. Juanes, Alberto Labarga, Antonio Fabregat, Vicente Arnau, Javier Chaves-

Martinez, Javier Garcia-Planells and Ana Barbara Garcia-Garcia

NGS facilities dedicated for clinical genomics need high QC standards, and they need continuously keeping track of their experiments and theirmetrics. As part of the Medical Genomics Visualization toolset (MGviz) we have developed an interactive software suit with R-Shiny and Python(Bokeh, crossfilter, flask and ReportLab) for automatic reports of QC for the whole NGS experiments in clinical diagnostics labs. The tool allowscomparison of the current experiment with historic data to see the performance of the sequencer, checking different metrics for coverage andvariations, warning for large copy number regions. In addition it remembers decisions over annotations, helps in variant prioritization,segregation and finding compound heterozygotes and make automatic historical reports of pathogenic variants informed by the lab.

NGS QC: ENAC accreditation in Process

- Automatic report after pipeline finishes.- Interactive coverage exploration.- All information of the NGS QC metrics stored in Mongodb and sqlite

Finding CNVs in Cancer

Control Discordant  profile

QC from historical Data

Check Sex

Find CNVs

Report coverage and gaps

DNA  agarose

Degraded

Amplicons

Exome

Gaps from historic p75

Gaps cov <20 Area historic percentil25-75Cov sample Area historic percentil25-75

Historical coverage per run

Comparing  against  historic  profiles  we  can  detect  degraded  DNA  or  other  artifacts

SummaryMGviz is a web and CLI tool very useful for QC andanalytics for exome and amplicon NGS. Here we haveshown only the QC and visualization tools.

Code  will  be  released  on  July  at  github.  https://github.com/MGviz-­‐soft  

pablo.marin@uv.es Doi:  10.6084/m9.figshare.3369787

Tumor Blood