Can I adding (multiply ) the P-value of one Transcriptioin factor identified by two motif?

Hi, I am using Homer to identify motif in a set of genomic regions, then I get a list of motifs with a list of best-matched transcription factors(TF). most of the time I get different TF from different motifs. It occurs to me this time that two motifs point to the same TF. I am wondering, should I just pick the one with the lower P-value, or should I multiply two p_value for this one TF?

Thanks!