RMG finds lots of unrealistic reactions in polyaromatics

[agvandeputte]

RMG overestimates the intra R additions in polycyclic compounds.

For example, in the indenyl radical RMG tends to find a whole plethora of Intra_R_Add_Endo/Exocyclic reactions that should not be occuring due to the high strain in the transition state.

[agvandeputte]

Issue resolved in be2d356 by implementing a check that only adds the intramolecular addition reaction if the addition site (double/triple bond) and the radical site are part of two adjacent rings (adjacent = the rings have two nodes in common)

Graph.java
    public boolean strainedRing(Node node1,Node node2) {
    boolean strained = false;
    for (int i = 0; i < SSSRings.size(); i++) {
        LinkedList cycle1 = (LinkedList) SSSRings.get(i);
        if (cycle1.contains(node1)) {
            for (int j = 0; j < SSSRings.size(); j++) {
            int atomCount = 0;
            LinkedList cycle2 = (LinkedList) SSSRings.get(j);
            if (i!=j && cycle2.contains(node2)) {
                for (int k = 0; k < cycle2.size(); k++) {
                    GraphComponent gc = (GraphComponent) cycle2.get(k);
                    if (gc instanceof Node) {
                       Node n = (Node) gc;
                       if(cycle1.contains(n))
                           atomCount += 1;
                       }
                    }
                }
            if(atomCount >= 2)
                strained = true;
            }
        }
    }
    return strained;
    }