The invention is directed to methods of generating sequence profiles of populations of nucleic acids, whose member nucleic acids contain regions of high variability, such as populations of nucleic acids encoding T cell receptors or B cell receptors. In one aspect, the invention provides pluralities of sets of primers for generating nested sets of templates from nucleic acids in such populations, thereby insuring the production of at least one template from which sequence reads are generated, despite such variability, or dispite limited lengths or quality of sequence reads. In another aspect, members of such populations are bidirectionally sequenced so that further sequence information is obtained by analyzing overlapping sequence reads in the zones of highest variability.