Go to content


- Estimating the local false discovery rate

An R/Bioconductor package for two-condition microarray experiments. Includes computation of scores, empirical p-values and local false discovery rates.


twilight logo twilight is an R/Bioconductor package that includes several functions for the statistical analysis of two-condition microarray data. The main algorithm for estimating the local false discovery rate is described in Scheid and Spang (2004a). The newer versions of twilight include the permutation filtering algorithm introduced in Scheid and Spang (2006).


  • Compensating for unknown confounders in microarray data analysis using filtered permutations Scheid S and Spang R.
    Journal of Computational Biology, 2007 Jun;14(5):669-81.
    pmid: 17683267  |  doi: 10.1089/cmb.2007.R009 ]
  • A stochastic downhill search algorithm for estimating the local false discovery rate
    Scheid S and Spang R.
    IEEE Transactions on Computational Biology and Bioinformatics 1(3):98-108.
    abstract  ]
  • Estimation of Local False Discovery Rates - User's Guide to the Bioconductor Package TWILIGHT Scheid S and Spang R.
    CompDiag Technical Report Nr. 2004/01 Nov. 2004

For further information please visit the compdiag projects Local False Discovery Rate or Permutation Filtering.


Download the package source or windows binary directly from the Bioconductor WWW-site. To run the package, you need R or higher and the R packages splines and stats. Package twilight includes a user's manual which can be downloaded separately from above.

→ the released version
→ the development version

Non-academic users

For commercial use of TWILIGHT, please visit the webpage of the NGFN Technology Transfer Center on genome-marketplace.

  1. Department of Medicine
  2. Institute of Functional Genomics