Instrumental setup and data reduction ALMA setup. ALMA Band 7 observations of HD142527 were carried out in the night of June 2 2012. The precipitable water vapor in the atmosphere was stable between 1.4 and 1.8 mm, with clear sky conditions. The ALMA correlator was configured in the Frequency Division Mode (FDM) to provide 468.750 MHz bandwidth in four different spectral windows at 122.07 kHz resolution (0.1 km/s) per channel. Each spectral window was positioned in order to target the CO(3-2) transition at 345.7959 GHz, HCO+ as well as CS(7-6) and HCN(4-3). The measured system temperatures ranged from 207 to 285 K in the different spectral windows. The number of 12 m antennas available at the time of the observation was 19, although two antennas reported very large system temperatures (DA41 and DV12) and were flagged during data reduction. Excluding calibration overheads, a total time on source of 52 minutes was spent yielding an RMS of 15 mJy in 0.1 km s −1 channels. The primary flux calibrator was Titan, which provided a mean transferred flux of 14.2 Jy for 3c279, the bandpass calibrator, and 0.55 Jy for J1604-446, the phase calibrator. Amplitude calibration used the CASA Butler-JPL-Horizons 2010 model for Titan, which gives an estimated systematic flux uncertainty of ⇠10%. All the line data were processed with continuum subtraction in the visibility domain. Image synthesis. Image synthesis was performed using two different techniques, depending on the application. For a traditional way to present the visibility dataset we use Cotton-Schwab CLEAN in the CASA package. This technique represents the consensus in image synthesis. We use Briggs weighting with robustness parameter of zero. For deconvolved models we use a non-parametric least-squares modeling technique 31 with a regularizing entropy term (i.e. as in the family of maximum entropy methods, MEM here and elsewhere). MEM model images are restored by convolving with the clean beam and by adding the residuals calculated using the difmap package 32. For the residuals we use weights comparable to our choice in CASA, a mixture of natural and uniform weights. A detailed example of this MEM algorithm is shown in the HCO + channel maps, Fig. S4. 1 Registration of ALMA images. A ⇠0.1 arcsec astrometric uncertainty could affect the ALMA data. However, we checked the astrometry by confirming that the centroid of the Keplerian velocity field (seen in the RGB image for CO(3-2) in Fig. 1) lies indeed at the po

We present 3D hydrodynamical models of the HD142527 protoplanetary disc, a bright and well studied disc that shows spirals and shadows in scattered light around a 100 au gas cavity, a large horseshoe dust structure in mm continuum emission, together with mysterious fast radial flows and streamers seen in gas kinematics. By considering several possible orbits consistent with the observed arc, we show that all of the main observational features can be explained by one mechanism -the interaction between the disc and the observed binary companion. We find that the spirals, shadows and horseshoe are only produced in the correct position angles by a companion on an inclined and eccentric orbit approaching periastron -the 'red' family from Lacour et al. (2016). Dust-gas simulations show radial and azimuthal concentration of dust around the cavity, consistent with the observed horseshoe. The success of this model in the HD142527 disc suggests other mm-bright transition discs showing cavities, spirals and dust asymmetries may also be explained by the interaction with central companions.