The information content (relative entropy) of transcription factor binding sites (TFBS) is used to classify the transcription factors (TFs). The TF classes are clustered based on the TFBS clustering using information content. Any TF belonging to the TF class cluster has a chance of binding to any TFBS of the clustered group. Thus, out of the 41 TFBS (in humans), perhaps only 5 -10 TFs may be actually needed and in case of mouse instead of 13 TFs, we may have actually 5 or so TFs. The JASPAR database of TFBS are used in this study. The experimental data on TFs of specific gene expression from TRRD database is also coinciding with our computational results. This gives us a new way to look at the protein classification- not based on their structure or function but by the nature of their TFBS.