Diagnostic Signature Challenge - Overview

Diagnostic Signature Challenge - Overview

The Diagnostic Signature Challenge is closed - keep testing your methods with the Diagnostic Signature Benchmarking Activity

The goal of the Diagnostic Signature Challenge was to assess and verify computational approaches that classify clinical samples based on transcriptomics data.

54 teams took part in the challenge between 5 March and 21 June 2012. Now you have the opportunity to compare your predictions to the actual results from the DSC by using the Diagnostic Signature Benchmarking Activity:

  • Register or log in to join the Benchmarking Activity of the Diagnostic Signature Challenge and read the detailed description of each sub-challenge
  • Download test data - Test data, alongside clinical information is available which does not reveal the test subjects diagnosis
  • Submit your prediction - You can upload your prediction file which will be instantly scored
  • Instantly download your results - See how you’ve performed against teams who completed the Diagnosis Signature Challenge

Please register or log in to join the Diagnostic Signature Benchmarking activity of the Diagnostic Signature Challenge

Please note that the use of this activity is subject to your agreement to the Activity Rules, the “Privacy Policy”and the Terms of Use”.

The purpose of the Diagnostic Signature Benchmarking activity is educational only. It aims to set a standard for Diagnostic Signature models, allowing scientists to objectively benchmark their methods. The tool automatically compares predictions to the results of the sbv IMPROVER Diagnostic Signature Challenge, which was run in 2012. The outcome of the Diagnostic Signature Benchmarking tool does not entitle participants to claim any prize or formal recognition. Your participation in this activity is governed by the Terms of Use, in particular Section 12 of the Terms of Use. You should carefully review the applicable terms and conditions to ensure you agree with the obligations and grants of rights contained therein. In particular, we note that:

  • the activity is not a challenge and there are no incentives or monetary awards available;
  • your use of the rankings provided by the activity is limited to internal, self-assessment only and do not constitute a endorsement by us;
  • you give us the right to publish your ranking and submission in aggregate with those of other participants in this activity and/or those of the participants in the Diagnostic Signature Challenge; and
  • you may use the data provided for the activity only for your participation in the activity.

If you wish to use the results of this tool in a scientific publication, we kindly ask you to cite us with following reference: [Hoeng, J., et al. (2013). sbv IMPROVER Diagnostic Signature Challenge: Preface to this special issue. Systems Biomedicine, 1] and inform us of your publication, which we will reference in our Publication Page


View Final Rankings of the Diagnostic Signature Challenge:

The final rankings are available here.

For more details about the Diagnostic Signature Challenge visit the "The Challenge" tab.

The best performing teams were announced at the sbv IMPROVER Symposium 2012 and the identity of the overall best performing team was published in Nature (Jan. 2013).

As published in Nature, 24 Jan. 2013, page 565


Share this page