Context. On March 28, 2011, the BAT instrument on board the Swift satellite detected a new transient event that in the very beginning was classified as a gamma ray burst (GRB). However, the unusual X-ray flaring activity observed from a few hours up to days after the onset of the event made a different nature seem to be more likely. The long-lasting activity in the X-ray band, followed by a delayed brightening of the source in infrared and radio activity, suggested that it is better interpreted as a tidal disruption event that triggered a dormant black hole in the nucleus of the host galaxy and generated an outflowing jet of relativistic matter. Aims: Detecting a very high energy emission component from such a peculiar object would be enable us to constrain the dynamic of the emission processes and the jet model by providing information on the Doppler factor of the relativistic ejecta . Methods: The MAGIC telescopes observed the peculiar source Swift J1644+57 during the flaring phase, searching for gamma-ray emission at very-high energy (VHE, E > 100 GeV), starting observations nearly 2.5 days after the trigger time. MAGIC collected a total of 28 h of data during 12 nights. The source was observed in wobble mode during dark time at a mean zenith angle of 35°. Data were reduced using a new image-cleaning algorithm, the so-called sum-cleaning, which guarantees a better noise suppression and a lower energy threshold than the standard analysis procedure. Results: No clear evidence for emission above the energy threshold of 100 GeV was found. MAGIC observations permit one to constrain the emission from the source down to 100 GeV, which favors models that explain the observed lower energy variable emission. Data analysis of simultaneous observations from AGILE, Fermi and VERITAS also provide negative detection, which additionally constrain the self-Compton emission component.