Genome-wide identification of transcription factor binding sites and gene regulatory elements is an important problem of computational genomics. Advances in high-throughput sequencing technologies combined with chromatin immunoprecipitation, such as ChIP-on-chip, ChIP-seq and ChIPPET (Paired-End diTag) allow us to map transcription factor binding sites (TFBS) and analyze mechanisms of gene regulation on the level of the entire genome. Examples include Oct4, Sox2, Nanog and 10 other transcription factors in mouse, and p53, c-Myc, ER, FoxA1 in human. Clustering of multiple binding sites by different TF reveals potential enhancer regions in mammalian genome. We discuss here statistical analysis of data mapping quality, sensitivity and specificity issues of the ChIP-seq TFBS sets and downstream gene expression analysis.