Aftershocks commonly bunch up near the ends of an initial earthquake rupture, in this case suggesting that yesterday's M7.7 broke about 200 km of the plate boundary. Statistically, if another significant quake occurs, it is most likely to be near the most active aftershock zone. 
That's what happened in Nepal in 2015: The Dolakha aftershock occurred near where the Gorkha rupture ended to the east, in an area that lit up with aftershocks right after the mainshock.
Never any guarantees in the earthquake business, of course, but the tendency for aftershocks to cluster allows seismologist to make statistical forecasts of both the expected number of aftershocks and their likely spatial distribution.
Only the first half of this, forecasting expected rates, is now implemented in USGS automatic aftershock forecasts. The spatial part is still a work in progress.
