All positional weight matrices from TRANSFAC (Matys et al. 2006), JASPAR CORE (Bryne et al. 2008), and UniPROBE (Berger et al. 2006, 2008; Badis et al. 2009; Newburger and Bulyk 2009; Scharer et al. 2009; Wei et al. 2010) were used to call motif matches in the genome. The PWMs were scored with a 1% added pseudocount and a uniform background. The TFM P-value program was used to determine a threshold at 4 × 10−8 for each PWM (Touzet and Varré 2007). All possible kmers that are above this threshold are aligned to the genome using Bowtie to give a final motif mapping (Langmead et al. 2009).
