際際滷

際際滷Share a Scribd company logo

GCAT Update June 2013 @ The Clinical Genome Conference

D
David Mittelman

GCAT (Genome Comparison and Analytic Testing) is a free online platform for benchmarking NGS methods and developing standards and metrics. It allows users to process sequencing data through different analysis tools and compare results. Since launching in April 2013, GCAT has been viewed over 20,000 times and has processed large amounts of sequencing data. New features continue to be added, including comparing variant calls to validation datasets and support for additional sequencing applications like RNAseq and de novo assembly. The goal is to accelerate adoption of NGS technologies by providing a common system for experimentation and validation of analysis methods.

1 of 16
GCAT: Genome Comparison and Analytic Testing Update June 2013 David Mittelman, Ph.D. Virginia Bioinformatics Institute at Virginia Tech @evolvability #TCGC #GCAT
How do we build the most accurate model or representation of a patients genome? Question:
This was supposed to be the future... where is my $1000 genome? Hall, N., Genome Biology. 2013.
Lots of platforms and sequencing applications
Even more ways to analyze the data Fonseca, N.A., et. al., Bioinformatics. 2012.
What standards do we use? How do we measure performance? http://xkcd.com/927/
http://xkcd.com/927/ @parkertm What standards do we use? How do we measure performance?
http://www.bioplanet.com/gcat Free to use for everyone A platform to benchmark NGS methods A collaboration effort for standards and metrics A common playground for experimentation An effort to accelerate NGS adoption A work in progress we need your help What is GCAT?
http://www.bioplanet.com/gcat What is GCAT?
Mapping algorithms continue to improve read alignments 100bp PE
Mapping quality performance is an important feature 100bp PE
The performance of mappers affect which variants are called
The performance of mappers affects variants discovery Mutation Recall: Genome in a Bottle Zook, J., et. al, in preparation. 2013.
Variant discovery is also greatly affected by variant callers Mutation Recall: Genome in a Bottle Zook, J., et. al, in preparation. 2013.
We launched GCAT to standardize genome analysis Launched April 7, 2013 at the Bio-IT World Conference. GCAT has been viewed more than 20,000 times from visitors across 144 countries. Our users have processed 500+ reports and more than 900 million bases. The GCAT cloud has processed 24 million variant calls and 1.6 billion alignments. All the data in GCAT is freely available for realtime 鍖ltering, comparing and sharing. http://www.bioplanet.com/gcat
New features in development for GCAT Lots of new features already available since our last presentation at Bio-IT World. For example data sharing via social media and advanced data 鍖ltering options. We will be imminently releasing new functionality to compare variant calls against the NIST Genome-in-a-Bottle truth set. Lots more validation data to come from Illuminas Platinum Genome Project, as well as orthogonal sequencing methods such as PacBio and Sanger. We are building support for somatic variant detection, whole genome sequencing, and RNAseq. New testing category: de novo assembly. http://www.bioplanet.com/gcat

More Related Content

GCAT Update June 2013 @ The Clinical Genome Conference

  • 1. GCAT: Genome Comparison and Analytic Testing Update June 2013 David Mittelman, Ph.D. Virginia Bioinformatics Institute at Virginia Tech @evolvability #TCGC #GCAT
  • 2. How do we build the most accurate model or representation of a patients genome? Question:
  • 3. This was supposed to be the future... where is my $1000 genome? Hall, N., Genome Biology. 2013.
  • 4. Lots of platforms and sequencing applications
  • 5. Even more ways to analyze the data Fonseca, N.A., et. al., Bioinformatics. 2012.
  • 6. What standards do we use? How do we measure performance? http://xkcd.com/927/
  • 7. http://xkcd.com/927/ @parkertm What standards do we use? How do we measure performance?
  • 8. http://www.bioplanet.com/gcat Free to use for everyone A platform to benchmark NGS methods A collaboration effort for standards and metrics A common playground for experimentation An effort to accelerate NGS adoption A work in progress we need your help What is GCAT?
  • 10. Mapping algorithms continue to improve read alignments 100bp PE
  • 11. Mapping quality performance is an important feature 100bp PE
  • 12. The performance of mappers affect which variants are called
  • 13. The performance of mappers affects variants discovery Mutation Recall: Genome in a Bottle Zook, J., et. al, in preparation. 2013.
  • 14. Variant discovery is also greatly affected by variant callers Mutation Recall: Genome in a Bottle Zook, J., et. al, in preparation. 2013.
  • 15. We launched GCAT to standardize genome analysis Launched April 7, 2013 at the Bio-IT World Conference. GCAT has been viewed more than 20,000 times from visitors across 144 countries. Our users have processed 500+ reports and more than 900 million bases. The GCAT cloud has processed 24 million variant calls and 1.6 billion alignments. All the data in GCAT is freely available for realtime 鍖ltering, comparing and sharing. http://www.bioplanet.com/gcat
  • 16. New features in development for GCAT Lots of new features already available since our last presentation at Bio-IT World. For example data sharing via social media and advanced data 鍖ltering options. We will be imminently releasing new functionality to compare variant calls against the NIST Genome-in-a-Bottle truth set. Lots more validation data to come from Illuminas Platinum Genome Project, as well as orthogonal sequencing methods such as PacBio and Sanger. We are building support for somatic variant detection, whole genome sequencing, and RNAseq. New testing category: de novo assembly. http://www.bioplanet.com/gcat