Recently, while analyzing a representative set of RNA structures from the PDB, I came across three weird entries. They are documented below, primarily for my own record.
- 5els — “Structure of the KH domain of T-STAR in complex with AAAUAA RNA”. There are two alternative conformations for the six-nt
AAAUAARNA component, labeled A and B, respectively. Normally, the A/B alternative coordinates for each atom are put directly next to each other, and assigned the same chain id, as in 1msy for the phosphate group of G2669 on chain A. In 5els, however, the two alternative conformations (A/B) are separated into two chains: chain H for A, and chain I for B.
- 1vql — “The structure of the transition state analogue ‘DCSN’ bound to the large ribosomal subunit of Haloarcula marismortui”. The three-nt fragment DA179—C180—C181 on chain 4 is in the 3’—>5’ direction.
- 4r3i — “The crystal structure of m(6)A RNA with the YTHDC1 YTH domain”. The mmCIF file has a model number of 0, instead of 1 (as in other cases I am aware of).