I've just disabled "Conservation filter" at
http://biomed.org.ua/COTRASIF/
to introduce necessary changes after E!54 release.
Also, public genomewide binaries were updated to tolerate non-Ensembl
fasta. However, this is only a quick hack, and "tolerate" is a good
description of it: whenever fasta record has a pipe "|" in it, it is
treated as non-Ensembl; binding sites are found OK, but coordinates
might be totally wrong (no verification has been performed yet, as
this is not an official feature).
I have used these new binaries to calculate some statistics on the
TFBS found in 2nd and 3rd exon sequences with zero cutoff - this
application does not need correct coordinates.
As time permits, support for non-Ensembl fasta will be improved to the
usable level.