Figure 2

Conversion process from DNA to SMILES. DNA is converted to amino acid sequences, which is compared using BLAST to find clusters of catalytic domains from secondary metabolic pathways. These domains are further analyzed to find the signature sequences, or specificity codes. These codes are compared using BLAST with an internal database of signature sequences to determine the individual polyketide or peptide activated by each catalytic domain. The individual SMILES of these polyketides or peptides are concatenated to construct the basic linear molecule as a SMILES string. Additional SMILES are produced from modifications to this basic SMILES set.