For a while, it seemed like we had more computationally predicted sequence motifs without a known matching transcription factor, than transcription factors without a known binding sequence ...