a, Surface diagram of FluB1 structure coloured as in c except that PA-C, PB1and PB2-N are uniformly green, cyan and red, respectively. The bottom black arrow indicates the extra 12 C-terminal residues of FluB PA that extend the PA C-terminal helix compared to FluA, so that it directly contacts the PB2-NLS domain that is consequently orientated slightly differently from in FluA polymerase. b, As in a but for bat FluA structure. Arrows highlight the 70° difference in orientation of the cap-binding domain. The structural similarity between FluA and FluB polymerases (LSQMAN, cut-off 3.5 Å) is as follows. PA: 630 Cα atoms aligned, of which 38.6% are identical with root mean squared deviation (r.m.s.d.) 1.34 Å; PB1: 703 Cα atoms aligned, of which 61.3% are identical with r.m.s.d. 1.06 Å; PB2: 428 Cα atoms aligned, of which 40.6% are identical with r.m.s.d. 1.46 Å (excluding the cap-binding domain), and, taking into account the cap-binding domain rotation, 622 Cα atoms aligned, of which 39.0% are identical with r.m.s.d. 1.54 Å). c, Subunit domain structure of influenza B polymerase with names and extended colour scheme, showing the positions of the PB1 polymerase motifs. Note that for PB1, the FluB numbering compared to FluA is the same from 1–399 and is thereafter +1. For PB2, FluB is +2 from 1–469 and +1 from 470–628. For PA it is more complicated owing to several short insertions and deletions. See Supplementary Fig. 1.