The signal elements in promoter sequences are not well characterized. We developed statistical tests to find nucleotide words (generally of length 8) that appear localized relative to TSSs (transcription start site). These words constituted """"""""seeds"""""""" for expansion to develop PSSMs (position-specific scoring matrices) characterizing systems of co-regulated genes. To this end, Dr. Marino-Ramirez collected a database of about 4700 sequences around the TSS of human genes, later increasing the size of the database by about a factor of 2. The database was exceptionally well characterized, and ideal for our statistical study. We used a Poisson scan statistic to determine whether occurrences of a given 8-letter DNA word are clustered unusually relative to the TSS. The Poisson scan statistic also identified clusters of significant words. About 80 of these words occurred in two or three clusters. By validating our results with microarray data and gene ontology information, we were therefore able to show that the same 8-letter word could have two different biological functions, depending on its position with respect to the TSS. Although this kind of positional dependency is a known phenomenon, our study showed that it is widespread in the human genome. In addition, with gold standard datasets and rigorous statistical tests, we showed that Markov models and positional information improve transcription factor binding site (TFBS) prediction significantly (although not yet to practical accuracies). Moreover, we showed that the Markov models used in extant TFBS programs is inferior, both theoretically and practically, to the theoretically correct Markov model we proposed. Positional information and the theoretically sound Markov models have been incorporated into the publicly available motif-finding program A-GLAM. Tatiana Orlova (Volunteer Jun-Jul 2009) and Narayan Perumal (Visitor Jul 2009) collaborated in using AGLAM to investigate possible TFBSs for the TOL-like receptors important to the immune response. John Spouge is also developing a computer program to extend the statistical methods to the known TFBS motifs in the JASPAR database, to categorize the motifs according to positional preference, and possibly to use positional preference to discover new instances of the TFBSs.
