Sequence Feature

Rules for marking up regions

• If a region is necessary and sufficient to drive a reporter gene, then mark it as an 'enhancer' or 'silencer'.

• If a region is both an enhancer and a silencer, then it should have the SO_term tags for both of these.

• If mobility shift experiments or similar experimental evidence is available to assert that a short region is a TF binding site, then mark it as a TF_binding_site.

• Similarity to a known binding motif is not evidence of being a TF_binding_site.

• If there is no evidence for a TF binding site and it has an effect on expression when mutated or deleted, but is not sufficient to drive a reporter gene, then we cannot assert that it is an enhancer or a TF binding site. Mark it as an anonymous 'regulatory_region'.

• If a region has the properties of being both a TF binding site and an enhancer then mark it up as two Features, one a TF_binding_site and one an enhancer.

• If a region is asserted to be a promoter region in the paper and it is within 200bp (or thereabouts?) of the 5' of the target gene and it is neccessary and sufficient to promote a reporter gene, mark it as a promoter. If in doubt, consider marking it as an enhancer.