Latest versions of DSSAT have interaction among N, P and K, but not well validated, needs calibration/validation with multi-location field trials.
QUEFTS deal with N, P and K optimization for target yield in a specific location.
Most of the field studies have not dealt with multi-nutrients interation for assessing growth and yield of crops/cropping system
There is a need to develop simple regression model on this aspect.
At this stage, my concern is any of the research working group working on this aspect.
For countries like India, the concern of yield stagnation could be answered through development of customized fertilizers (all nutrients in one for a target yield in a specific growing environment)