Intrinsic correlations among nucleosome positioning, presence/absence of TATA box and 5'UTR length. (A) Average 5'UTR lengths among the four gene groups clustered by k-mean clustering based on their nucleosome occupancy patterns. The mean value of each group is indicated by a bar and the error bars indicate one standard error. (B) The nucleosome occupancy in the 1201 bp window surrounding the TSS in the three groups of genes with different 5'UTR lengths. The TSS proximal region of long 5'UTR genes has the highest nucleosome occupancy. (C) The proportion of TATA box-containing genes and TATA box-less genes in each 5'UTR length group of genes. The long 5'UTR genes have the highest proportion of TATA box-containing genes, but the lowest proportion of TATA box-less genes. (D) TATA box-containing genes have slightly longer 5'UTRs than do TATA box-less genes.