mod(mdg4) all 3 adjacent but CG15802 on opposite strand This is because part of the Celera sequence is in the wrong orientation WRT surrounding sequence. Technically, this is a merge, but do we want to merge CGs on opposite strands? > mod(mdg4) FBgn0002781 = CG15802|FBan0015802|CT40910|CT20301|CT40908| AND CG7836|FBan0007836|CT23629| AND CG8076|FBan0008076|CT23762| The problem is in the region around 136941-138566 of AE003734 (these are approximate coordinates based on cDNA alignment, I am not sure where the exact flip is). I think it is the Celera sequence that's wrong and not the cDNAs in GenBank because the cDNAs from different labs agree with one another.