Features of a number of important but poorly explained human clinical
syndromes strongly indicate a microbial etiology. In these syndromes,
the failure of cultivation-dependent microbial detection methods revea
ls our ignorance of microbial growth requirements. Sequence-based mole
cular methods, however, offer alternative approaches for microbial ide
ntification directly from host specimens found in the setting of unexp
lained acute illnesses, chronic inflammatory disease, and from anatomi
c sites that contain commensal microflora. The rapid expansion of geno
me sequence databases and advances in biotechnology present opportunit
ies and challenges: identification of consensus sequences from which r
eliable, specific phylogenetic information can be inferred for all tax
onomic groups of pathogens, broad-range pathogen identification on the
basis of virulence-associated gene families, and use of host gene exp
ression response profiles as specific signatures of microbial infectio
n.