Download
quality control of thousands of experiments with qcml n.
Skip this Video
Loading SlideShow in 5 Seconds..
quality control of thousands of experiments with qcML PowerPoint Presentation
Download Presentation
quality control of thousands of experiments with qcML

quality control of thousands of experiments with qcML

51 Views Download Presentation
Download Presentation

quality control of thousands of experiments with qcML

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. quality control of thousands of experiments with qcML pieter kelchtermans pieter.kelchtermans@vib-ugent.be computational omics and systems biology group VIB / Ghent University, Ghent, Belgium

  2. How do you want to do quality control What we tried to look at and how we did it What we're looking forward to

  3. We automated OpenMS Process 5098 raw files to output qcML Using Mathias Walzer’s workflow Completely automated, on 11 linux computers

  4. Preliminary analysis shows

  5. You choose how to control the quality

  6. qcML is a universal format Captures metrics and metadata about all kinds of experiments Separate from the actual analysis, only a container Metrics are added with controlled vocabulary

  7. qcMLoffers fast and easy access QCExporter/QCExtractor gets data in tabular format Java library under development by WoutBittremieux R integration with library by Laurent Gatto

  8. Metrics can be generated by OpenMS QuaMeter Simpatiqco You

  9. Large-scale is not intractable Automated deployment of virtual machines OpenMS updates/installation Configuration Batch data processing Shared folder between all computers?

  10. Be the master of your puppets http://www.puppetlabs.com/ since 2005 Open source, cross platform agent package { “openms”: ensure => "installed" } file { “/etc/openms.ini”: source => “puppet:///modules/openms/openms.ini“ }

  11. Each kind of experiment has different quality thresholds

  12. Experiments vary a lot Classification between experiments Boundaries need to be determined by large-scale analysis qcML allows for thresholds: flag bad experiments

  13. Ratio of +1/+2 charged features

  14. Ratio of +3/+2 charged features

  15. Mean precursor m/z 900 600

  16. Total Ion Current jumps 4000 3000 2000 1000

  17. We’re shipping by the end of this year Format definition: http://code.google.com/p/qcml/ Advanced analyses on differences between runs Get in touch!

  18. WoutBittremieux Mathias Walzer Jonathan Vandenbussche, Kris Gevaertlab www.compomics.com @compomics