Enhancer Element Locator

Enhancer Element Locator, or EEL, is a tool for locating distal gene enhancer elements in mammalian genomes by comparative genomics.

This web page tries to provide some guidance in the usage of the EEL sofware. These instructions refer to the TclTK/Tix user interface as seen in the image below. This interface is available on Windows and on most Unix systems with Xwindowing environment. The important exception is the MacOS X which has it's own MacEEL.

You can try out EEL with example sequences and binding site matrices. Ruutukaappaus.png The basic procedure for comparing two orthologous sequences is simple as follows

  1. Add the orthologous sequences in FASTA formated files by clicking "Add Sequences".
  2. Add the binding site matrices in pfm formated files by clickin "Add Matrices". This format is used for example in JASPAR.
  3. Get the binding sites by clicking "Get Binding Sites" and "Get TFBS".
  4. Align the binding site sequences by clicking "Align Sites" and "Align".
  5. Look at the results by clicking "Show Alignments".
The two later steps take a whole bunch of parameters which are described in the article and its supplement. The defaults should be OK in general.

Removing a sequence or matrix from the user interface is done by double clicking its name.

You can look and save the list of putative binding sites (mostly junk) by clicking "Show sites" and the list of predicted enhancer elements by clicking "Show alignments". If the DNA sequence is added to the program, the alignment output will be nice "DNA alignment like". If not (you have either removed the sequences or aligned a stored list of putative binding sites), the output will be just numbers.

Output

EEL outputs either easily parsable GFF format or more human readable custom format as described below.

### lambda=2 mu=0.5 nu=200 xi=200 Nucleotides per rotation=10.4
### D[ENSMUSG00000037169][ENSG00000134323]
Note! First nucleotide at position 1 (one) and binding site at zero!
First lines give out the used parameter values and the names of the used sequences. Here the sequence names are ENSMUSG00000037169 and ENSG00000134323 
### Alignment No 1 ###
D[1762][2445]=38.72 S8.pfm (18430,18434) <=> (24317,24321) +
D[1766][2449]=75.43 Tal1beta-E47S.pfm (18462,18473) <=> (24349,24360) +
D[1772][2451]=92.13 Pax6.pfm (18527,18540) <=> (24414,24427) +
D[1774][2455]=111.26 Broadc4.pfm (18583,18593) <=> (24470,24480) -
D[1777][2456]=151.13 SOX17.pfm (18603,18611) <=> (24490,24498) +
D[1779][2461]=196.33 NF-kB.pfm (18620,18629) <=> (24507,24516) +
D[1780][2464]=232.21 Thing1-E47.pfm (18647,18656) <=> (24534,24543) -
D[1787][2470]=216.88 S8.pfm (18763,18767) <=> (24649,24653) -
D[1788][2471]=254.45 Myf.pfm (18776,18787) <=> (24662,24673) +
On the fist line is the rank of the alignment (the best local alignment on this region).
On the following lines: The values in the brackets are the indexes of the aligned binding sites in the sorted order of binding sites on either sequence. Equals the score of the alignment so far. Name of the aligned binding site. (start,end) positions on the sequences. Finally, the strand where the binding site is located. 
Sequence 1: ENSMUSG00000037169
Sequence 2: ENSG00000134323

  18421 : g-aagagaacAATTAagttc-ttttccctagg-catgtgtggaaTCAACATCTGGAtgcc
  24308 : gcaa-agagcAATTA-gcttcttttct-tgggacacatatagaaTCAACATCTGGA-gac

  18478 : t-cgagcctgaggcccattacaagtagcaaaagaatttggacaggatgcaTTTAATCTTG
  24364 : cacgagcctgagacccatttcaagtagcaaaagaatttggacaggatgcaTTTAATCTTG

  18537 : AGTTaatggtacagtgctccgccaaga-aaaacctctcagcttcaacATATTTTACTTag
  24424 : AGTTaatggtacaaagctcagccaagataaaat-tctcagcctcaacATATTTTATTTtg

  18596 : aaaaactGTTATTGTCaggtacagGGGAAATTCTtcctccctttggctgctCTGCCAGAT
  24483 : aaaaaatGTTATTGTCaggtacaaGGGAAATTCCtcctaacttagggggctCTGCCAGAT

  18656 : Gacatagttactgcaggggaccact-gacctggtgtgttatctctttcatctgagaaagt
  24543 : Gacataagcactggaggg-accattcgtc-tggtgtgttatctttttcatctaagtaggt

  18715 : ctctcccccgcagactcatctcccccaaacctggccatgctccctgtgTAATTctagcct
  24601 : ctttcctccaccaactcatctctcaaaaagctggccacagttcctaagTAATTctatccc

  18775 : cTGACAGCTGCAGgagaaggaag-
  24661 : cTGACAGCTGCAGga-aagaaaat
Finally there is the DNA like alignment. The EEL algorithm proper have aligned only the sequences with upper case letters. The lower case nucleotides are aligned just for illustrative purposes.

Terminal and Command line interfaces

You can use EEL also from text terminal or from command line. This mode provides more features to be used than the GUI mode and allows using EEL in non-interactive fasion for example in batch runs. If EEL can not find TclTK/Tix libraries, it will revert to the terminal mode and you can reach this mode by using command line parameter -no-gui

You can use eel on command line by giving terminal commands preceeded by '-' on the command line. The commands are executed from left to right.

In the terminal mode, command 'help' provides the list of available commands and their short introduction. 

'-?', '-h', '-help'
        Arguments: none
        prints this help

'-about'
        Arguments: none
        Prints Information about the program

'-addMatrix', '-am'
        Arguments: filelist
        reads matrices from files

'-addSequence', '-as'
        Arguments: filelist
        reads sequences from files

'-addSingleSequence', '-ass'
        Arguments: filelist
        Gzipped and Fasta formated sequence files. One sequence in file.

'-align'
        Arguments: [filename[,num_of_align,[lambda[,xi[,mu[,nu,[,nuc_per_rotation]]]]]]]
        aligns the computed BS or optional the BS from a gff file
        filename specifies a file in gff format is you want to be aligned
        num_of_align        specifies how many alignments you want. (Default 3)
        lambda   Bonus factor for hit (Default 2)
        xi       Penalty factor for rotation (Default 1.0)
        mu       Penalty factor for average distance between sites (Default 0.5)
        nu       Penalty factor for distance difference between sites (Default 1.0)
        nuc_per_rotation    specifies how many nucletides there are per rotation. (Default 10.4)
        If you want to skip a argument just  write '.' for it.
        If you use '.' as filename the local data are aligned.

'-cd'
        Arguments: [path]
        Change or display the current working directory.

'-dir'
        Arguments: [path/pattern]
        List files matching the given pattern. Defaults to "*" in current working directory.


'-getTFBS'
        Arguments: [bound]
        computes the scores of all matrices and all sequences which are
        better than bound*maxscore. maxscore is the highest reachable
        score of the actual matrix with respect to the background
        The default value for bound is 0.1

'-getTFBSabsolute'
        Arguments: [cutoff]
        computes the scores of all matrices and all sequences which are
        better than cutoff.
        The default value for cutoff is 9.0

'-more'
        Arguments: [number of alignments]
                prints more alignments from previously run alignment matrix

'-no-gui'
        Arguments: none
        Gives command line interface

'-pm', '-printMatrices'
        Arguments: none
        prints the matrices

'-pmw', '-printMatrixWeights'
        Arguments: none
        prints the matrix weights (with background)

'-printSeqNames', '-ps'
        Arguments: none
        prints the names of the sequences

'-q', '-quit'
        Arguments: none
        to exit the program

'-removeMatrix', '-rm'
        Arguments: Matrixnumber
        removes a matrix

'-removeSequence', '-rs'
        Arguments: Sequencename
        removes a sequence

'-reset'
        Arguments: none
        removes all matrices and sequences

'-resetMatrices', '-resm'
        Arguments: none
        removes all matrices

'-resetSequences', '-ress'
        Arguments: none
        removes all sequences

'-sa', '-showalign'
        Arguments: none
        prints the computed alignment to stdout

'-saveMarkovBackground'
        Arguments: filename
        Name of the file where to store the background model.


'-savealign'
        Arguments: [filename]
        saves the alignment to disk
        The default filename is 'eel_[Date+Time].align'
        e.g. eel_2003_9_16_11_48.align

'-savealignAnchor'
        Arguments: [filename]
        saves the alignment to disk in Anchor format for DIALIGN
        The default filename is 'eel_[Date+Time]_align.anc'
        e.g. eel_2003_9_16_11_48_align.anc

'-savealignGFF'
        Arguments: [filename]
        saves the alignment to disk in GFF format
        The default filename is 'eel_[Date+Time]_align.gff'
        e.g. eel_2003_9_16_11_48_align.gff

'-savematch'
        Arguments: [filename]
        saves the results of the matching in gff format
        See http://www.sanger.ac.uk/Software/formats/GFF/
        The default filename is 'eel_[Date+Time].gff'
        e.g. eel_2003_8_27_15_48.gff

'-setBGfreq'
        Arguments: A C G T
        Background nucleotide frequencies. Removes markov background.

'-setMarkovBG'
        Arguments: bgSampleSequence [order]
        Background sample sequence and order of the model or saved background file.

'-setpseudocount'
        Arguments: [pseudocount]
        Set the amount of pseudocounts on matricies. Default 1.0

'-showmatch', '-sm'
        Arguments: none
        prints the computed scores to stdout

References

">Kimmo Palin
Last modified: Mon Jan 16 14:13:32 EET 2006