RegulonDB
RegulonDB 9.4: Position specific scoring matrices for TFs - Help
Position specific scoring matrices for TFs - Help

TF Name:Transcription factor name

Matrix type: Name of the program used to make the matrix using the TFBSs from RegulonDB.

Width:
Size of the matrix built.

Background model:

(http://rsat.ulb.ac.be/help.matrix-quality.html)

An essential parameter of any pattern discovery or matching approach is the choice of an appropriate background model. This model is used to estimate the probability for each site to occur by chance given the sequence nucleotide composition, rather than as an instance of the motif.

The program matrix-scan supports Markov models of arbitrary order as background models. A Markov model of *order m* means that the probability of each residue depends on the *m* preceding residues in the sequence. Note that a Markov model of order 0 corresponds to a Bernoulli model (each residue is independent from the preceding ones).

The background model used for the evaluation was the upstream region with Markov order 0 and 1.

Logo

Sequence logo, a visual representation of the motif, where each column of the matrix is represented as a stack of letters whose size is proportional to the corresponding residue frequency. The total height of each column is proportional to its information content.

Sequence logo are generated using the freeware program Weblogo.

Score distributions logy

Decreasing cumulative distribution function (dCDF) (logy)

Abscissa represents the WS assigned by the matrix. Note that the Y-axis is in log-scale, in order to emphasize small frequencies.

ROC curves logX

The ordinate indicates the sensitivity (fraction of sites detected), the abscissa shows the corresponding FPR. Note the logarithmic X-axis, which is essential to highlight the relevant FPR range (small values)

ROC curves logX

The ordinate indicates the sensitivity (fraction of sites detected), the abscissa shows the corresponding FPR. Note the logarithmic X-axis, which is essential to highlight the relevant FPR range (small values)

Score distributions

Decreasing cumulative distribution function (dCDF)

Info matrix quality

It is a link to display more details of the results of the matrix-quality program.

For details of the matrix quality program and parameters see http://rsat.ulb.ac.be/help.matrix-quality.html

RegulonDB