In the paper , the following talk about the degenerative nature of "ChIP-seq" dataset. After a while googling, I still can not understand what is the meaning of "degenerative nature". Does it mean: Only part of the binding site were represented by the PWMs, there are still a large number of binding sites can not be detected by ChIP-Seq, so the scanning based on the PWMs should give more new TFBS. Isn't it?
Although the PWMs were derived from the corresponding ChIP-seq dataset, due to their degenerative nature, we still expected to obtain PWM hits that did not overlap with the ChIP-seq peaks.
 Gong, W., et al. (2015). "Inferring dynamic gene regulatory networks in cardiac differentiation through the integration of multi-dimensional data." BMC Bioinformatics 16: 74.