I'm currently using DANPOS2 to profile nucleosome occupancy genome-wide using data from an MNase-seq experiment. We're interested in changes in nucleosome occupancy related to a specific transcription factor binding, comparing occupancy with agonist and vehicle.
DANPOS2 profiles at specific genomic features like Transcription Start Site, Transcription Termination Site, etc, all of which are provided by an annotation file of all known genes in the mouse genome from UCSC genome browser. This is great but it's genome-wide so I really don't see differences as we only expect to see changes in a handful of specific target genes. Using UCSC genome table browser tool I was able to make a list of specific genes, for example 20 known target genes and 20 genes known to not be targets. So the signal to noise ratio is much better than genome wide. However, it's still comparing nucleosome occupancy at the same genomic features that UCSC includes in the columns of the annotation file, TSS and so on. I'm most interested in using the conserved binding element for the transcription factor as the genomic feature so we are comparing occupancy at the transcription factor binding site where we believe changes in nucleosome occupancy are taking place. Is this possible?
Our lab as well as others have done ChIP-seq experiments IP'ing this transcription factor, so the known binding sites are documented. Could the BAM files from these ChIP-seq experiments be used to create an annotation file? Or is there a repository with these tracks/annotations already available in BED format? I'm not sure if what I'm asking is really even possible or if the DANPOS2 program would even accept such a file if the columns don't match up with what it expects?
