Guillermo Barturen, Antonio Rueda, José L. Oliver and Michael Hackenberg
MethylExtract: High-Quality methylation maps and SNV calling from whole genome bisulfite sequencing data
How to cite

For questions and feedback, please contact to: MH -, JLO -, GB -


News and releases

  • Version 1.8 is available.

  • Version 1.7 is available.

  • Version 1.6 is available.

  • Version 1.5 is available.

  • Version 1.4 is available.

  • Version 1.3 is available.

  • Version 1.2 is available.

  • Version 1.1 is available.

  • Version 1.0 is available.




  • Version 1.8: 26/05/2014

    Solved a bug during the binomial coefficient calculation on (Thanks to Corey Wischmeyer for reporting the bug)


    Gapped alignment handling improved

    Checking parameters step removed

  • Version 1.7: 13/03/2014

    Solved a bug which affects to reads that go further than the reference assembly, this situation could appear when using seed alignment methods (Thanks to Dr. Huang for pointing out the bug and testing the developing version)

    Solved a sequence output bug on CHH contexts


    Included explanations to extract single chromosome reads and fasta sequences from multifasta files


    Output subroutines rewritten (easier to understand and faster)


    Parameters adapted


    Version removed from files names

  • Version 1.6: 06/03/2014

    SNVs Fisher strand bias threshold included

    3' nucleotides filter included

    Pair-End overlapping segments solution included


    Multifasta support included

    *.fas and *.fasta extensions are accepted as Fasta files


    No context coverage warning included (probably due to a Qscore or FLAGs erroneously selection)

    FLAGs for Watson and Crick strands are now mandatory parameters

    PARAMETER NAMES CHANGED (more informative names)

    tagW and tagC has been deprecated (new parameters flagW and flagC)

    chromSorted has been deprecated (new parameter chromSplitted)

  • Version 1.5: 13/02/2014
  • Log output files bug fixed

    To ensure directional consistency, the Fisher strand bias is defined as 1 minus the p-value (the higher is the new value, the higher the strand bias)

    Multiple FLAGs can be assigned to watson or crick mapped reads (tagW & tagC) - (Thanks to Dr. Krueger for pointing out this pair-end limitation)

  • Version 1.4: 04/11/2013
  • Fixed a bug related to perl versions below 5.14 -(Thanks to Dr. Stadler and Dr. Feehery for pointing out this bug)

    MethylExtract uses input BAM files without converting the entire file to SAM

  • Version 1.3: 10/11/2012
  • script extracts all detected SNVs (MethylExtractOutMethContexts has been included in the main script).

    The memory usage has been reduced.

    Default SNVs output is now VCF, check the manual or the output file header.

    Default Methylation output retrieves more information and has a more compressed format.

    Changed the SNVs calling (similar to varScan), giving more reliable results.

  • Version 1.2: 17/09/2012
  • Now the chromosome IDs will be taken from the interval fasta ID (solving compatibility bugs with Bismark aligner).

    Missing values in the output files will be filled up with '.' instead of '0'.

    Included a script to calculate the bisulfite conversion rate from an unmethylated genome (

    Included a script to calculate the error probability for each methylation value and assign a FDR to the entiry dataset (

    Included a script to extract SNVs and Methylation ratios out of reference methylation contexts (

  • Version 1.1: 19/07/2012
  • Included minimum depth parameters ("minDepth").

    Improved command line information and help.

    Increased log-file information.

  • Version 1.0: 20/05/2012
  • All basic functions working.






© Computational Genomics and Bioinformatics Group
Dept. of Genetics & Inst. of Biotechnology
University of Granada, Spain