Members of the genus Candidatus Accumulibacter are important in many wastewater treatment systems performing enhanced biological phosphorus removal (EBPR). The Accumulibacter lineage can be subdivided phylogenetically into multiple clades, and previous work showed that these clades are ecologically distinct. The complete genome of Candidatus Accumulibacter phosphatis strain UW-1, a member of Clade IIA, was previously sequenced. Here, we report a draft genome sequence of Candidatus Accumulibacter spp. strain UW-2, a member of Clade IA, assembled following shotgun metagenomic sequencing of laboratory-scale bioreactor sludge. We estimate the genome to be 80-90% complete. Although the two clades share 16S rRNA sequence identity of >98.0%, we observed a remarkable lack of synteny between the two genomes. We identified 2317 genes shared between the two genomes, with an average nucleotide identity (ANI) of 78.3%, and accounting for 49% of genes in the UW-1 genome. Unlike UW-1, the UW-2 genome seemed to lack genes for nitrogen fixation and carbon fixation. Despite these differences, metabolic genes essential for denitrification and EBPR, including carbon storage polymer and polyphosphate metabolism, were conserved in both genomes. The ANI from genes associated with EBPR was statistically higher than that from genes not associated with EBPR, indicating a high selective pressure in EBPR systems. Further, we identified genomic islands of foreign origins including a near-complete lysogenic phage in the Clade IA genome. Interestingly, Clade IA appeared to be more phage susceptible based on it containing only a single Clustered Regularly Interspaced Short Palindromic Repeats locus as compared with the two found in Clade IIA. Overall, the comparative analysis provided a genetic basis to understand physiological differences and ecological niches of Accumulibacter populations, and highlights the importance of diversity in maintaining system functional resilience.