Methods, computer systems, and computer readable medium for testing a plurality of models in order to classify a biological specimen. A determination is made as to whether a model precondition associated with a model in the plurality of models has been satisfied. The model is skipped when the model precondition has not been satisfied. When the model precondition has been satisfied, the first model is tested by selecting a calculation from the first model. The selected calculation is computed using cellular constituent abundance values from the biological specimen in accordance with a calculation algorithm set forth in the selected model. This selecting and computing is repeated for one or more calculations in the selected model. Then, each of these calculations is aggregated in accordance with an aggregation algorithm associated with the selected model. These steps are repeated for at least one other model in the plurality of models. A computer medium comprising a classifier for classifying a biological specimen into a biological sample class that includes one or more models. Each of the one or more models comprises (i) an optional model precondition, (ii) a calculation algorithm, (iii) a plurality of calculations, and (iv) a calculation aggregation algorithm.