NAME - Filter output files from by different criteria.

SYNOPSIS [options] "filter" [file]


-c, --count

Count the windows/loci instead of printing them.

-v, --version

Prints version information and exits.

-h, --help

Print a short help message and exits.


Prints a detailed manual page and exits.

DESCRIPTION reads tab-delimited data files as generated by For each window a filter is applied and if the filter is passed the window and the corresponding locus are printed out. Thus, you get all loci with at least one window that fulfills your filter criteria.

The mandatory filter statement is given within double quotes (" ") and can contain comparison/logical statements and field identifiers as listed below.

Technically, the statement is directly interpreted by Perl, so you can use anything which works in Perl. The same caveats apply, for example: If you want compare numbers you must use ==, if you compare strings you have to use eq.

Please note: everything you put in the filter statement is evaluated by Perl. This can be potentially harmful, so take care.


  1. windowID

    Consecutive numbered ID for each window

  2. locusID

    The locus which this window belongs to

  3. seqID

    Identifier of the sequence (e.g. human.chr1 or contig42)

  4. start

    Start position of the reference sequence in the window

  5. end

    End position of the reference sequence in the window

  6. strand

    Indicates if the reference sequence is from the positive or negative strand

  7. N

    Number of sequences in the alignment

  8. columns

    Number of columns in the alignment

  9. identity

    Mean pairwise identity of the alignment

  10. meanMFE

    Mean minimum free energy of the single sequences as calculated by the RNAfold algorithm

  11. consensusMFE

    ``Consensus MFE" for the alignment as calculated by the RNAalifold algorithm

  12. energyTerm

    Contribution to the consensus MFE which comes from the energy part of the RNAalifold algorithm

  13. covarianceTerm

    Contribution to the consensus MFE which comes from the covariance part of the RNAalifold algorithm

  14. combPerPair

    Number of different base combinations per predicted pair in the consensus seconary structure

  15. z

    Mean z-score of the sequences in the alignment

  16. SCI

    Structure conservation index for the alignment

  17. decValue

    Support vector machine decision value

  18. P

    RNA class probability as calculated by the SVM

  19. COL#

    Specify a particular column by its index #. First column has index 1. e.g. COL18>0.9 is equivalent to P>0.9



Less than, greater than


Equals numerically


Equals (strings)


Matches regular expression.

(, ), and, or, not

Logical operators and grouping


 # "P>0.9 and z<-3 and seqID~=/chr13/" results.dat

Gives you all clusters with windows with P>0.9 and z<-3 on chromosome 13.

 # -c "P>0.9" results.dat

Counts all windows/loci with P>0.9.


Stefan Washietl <>