New small-molecule drugs are needed, both to address existing disease and to combat the rise of antibiotic resistance in pathogenic microbes. A major source of antibiotics and other valuable therapeutic agents remains the natural products produced by Streptomyces and allied bacteria. The advent of genome-level sequencing has changed how such bioactive products are identified in two ways: first, by enabling new approaches to engineering of known clusters to obtain new analogues; and secondly by enabling mining of their large (8-12 Mbp) genomes to make rapid links between valuable compounds and the gene clusters responsible for their biosynthesis; as well as a complete inventory of orphan clusters for decryption. In this work, both of these aspects were explored. Accelerated Evolution (AE) is a new method of induced recombination in Streptomyces that produces libraries of analogues of a parent compound in a single experiment. It has been previously applied to a typical assembly-line biosynthetic pathway involving a multimodular polyketide synthase, that for the immunosuppressant rapamycin. Here, the utility of AE was initially explored with several polyketide antibiotics, before focussing on the polyene filipin. Unfortunately, this system proved refractory to the AE method, since only recombinants which had either lost filipin production, or which maintained some level of filipin production, were recovered. These results suggested that deeper understanding of the fundamental processes of recombination in Streptomyces are now needed to allow the AE approach to flourish. A genome mining approach was taken to identify and analyse the gene cluster in Streptomyces albus DMS40763 for pseudouridimycin (PUM), a rare C-nucleoside antibiotic and a newly-recognised selective inhibitor of bacterial RNA polymerase. The fully-sequenced genome (8.01 Mbp) of S. albus was analysed and 27 biosynthetic gene clusters were found, including those for 11 novel assembly-line systems. A strong candidate for the PUM gene cluster was identified through BLAST searches with a gene probe (truD) found in the gene cluster for a different C-nucleoside in other Streptomyces spp., hinting at a common mechanism for formation of the C-nucleoside moiety. Identification of the enzymes predicted to be encoded by the gene cluster allowed a detailed biosynthetic pathway to be proposed. Several genes from the PUM cluster were expressed and purified as recombinant proteins in Escherichia coli, and the proposed enzymatic roles of two of them were verified. One of these, a pseudouridine oxidase, also showed activity against uridine. Consistent with this finding, feeding of uracil to a mutant S. albus lacking the pseudouridine synthase-like gene truD gave production of the novel N-linked analogue of PUM, suggesting that the biosynthetic enzymes may be sufficiently flexible to incorporate other, non-natural nucleosides into PUM analogues for testing.