Position frequency matrices (PFM) from each source were converted to position probability matrices (PPM).
Transcription factor binding site (TFBS) motifs can be accurately represented by position frequency matrices (PFM) or other equivalent forms.
Position frequency matrices (PFMs) were constructed for Crx, Nrl, and Nr2e3 based on previous literature [10], [18], [56].
To identify potential binding sites for transcription factors, we searched through TRANSFAC database for position frequency matrices that match the motifs found by DME.
Position frequency matrices (PFMs), the most widely used models to represent binding specificities, are experimentally characterized only for a small fraction of all TFs.
In the following, these empirical substitution frequency matrices corresponding to 1 PAM will be simply referred to by their common acronyms, JTT, WAG, LG, KHG, mtREV, and cpREV.
We also used the program CLOVER [70] that uses the position frequency matrices (PFMs) of cis-regulatory sites to evaluate sequences for statistically significant over/under-representative sequence elements.
However, for genome-wide applications, i.e., the computational inference of transcriptional regulatory networks, more simple representations of DNA-binding specificities, such as position frequency matrices (PFMs) are used more commonly [27] [29].
Element-specific frequency matrices were generated from upstream sequences known to confer ABA responsiveness.
Here, we focus on DNA motifs represented by position frequency matrices (PFMs; Stormo, 2000).
Frequency matrices were constructed utilising a set of ABREs and CE3s obtained from the literature.
