Atlas for drug discovery.

Reports of increasing antibiotic resistance and sagging drug discovery rates are appearing with increasing frequency, and many warn that we are living on the wrong side of an antibiotic peak—a period in which everfewer new antibiotics are being discovered at ever-increasing costs (1). In response, efforts to discover new antibiotics and other drugs have taken many forms, including looking at formerly productive sources that were thought exhausted, much as rising oil prices led to new extraction techniques for abandoned oil fields. For several decades, small molecules produced by soil bacteria were our most important source of new drugs, as represented by the antibiotics erythromycin and vancomycin, the anticancer agents bleomycin and mitomycin, and the immunomodulators cyclosporin and rapamycin (2). In the last two decades, almost all pharmaceutical companies have abandoned bacterially based drug discovery because it seemingly fits poorly with the high-throughput screening and medicinal chemistry approaches that define the industry’s favored discovery paradigm, and because its high rate of rediscovering previously known compounds indicated that it was unlikely to yield new drugs (3). More recently, genomic revelations have dramatically altered our view of soil bacteria. The revelations were of two sorts: (i) most bacteria (∼99%) cannot be cultured under typical laboratory conditions, and (ii) even those that can be cultured produce only a fraction (∼10%) of the small molecules encoded in their genomes (4–6). These dual shortfalls in culturing and expression pointed to a bonanza of potentially useful small molecules remaining to be discovered and led to innovative technical approaches to finding them. In PNAS, Charlop-Powers et al. (7) use cultureindependent DNA sequencing on a broad environmental scale to systematize where different types of bacterially produced small molecules are likely to be found. Their “chemicalbiogeographic survey” provides insights into microbial ecology along with a practical guide for microbial genome prospectors. The potential scale of such a survey is staggering. Bacteria are the most diverse organisms on the planet, and a single gram of soil can contain between a thousand and a million distinct species according to a pioneering survey (8). Species definition in bacteria has its whimsical aspects, but it usually means that a characteristic DNA sequence from one strain has, or does not have, a specified similarity to the sequence from another strain. If the similarity is high enough, say 98%, then the two strains are said to belong to the