First draft attempt at some Q2Q causal relations.

[dosumis]

See See #350
DON'T MERGE YET! More discussion needed.

[dosumis]

Currently going with phenotype_results_in and phenotype_results_from, maybe simplify to phenotype results_from / results_in ?

There is demand for weaker sounding relations, do we add these?

Some examples (hemolytic anemia) might better be considered P->Q not Q->Q

For use within a phenotype def, should we restrict direction to results_from?

Ultimate aim is to define with reference to process -> process causal relations if poss but this will take work.

[matentzn]

Does the label "phenotype" capture the wild type (i.e. normal phenotypes) as well? If so, I think I am good with this.