Advancements in sequencing technology have made it possible to acquire large amounts of genomic data in a relatively cost-effective fashion. However, accurate and reliable genome assembly remains a challenge. The work presented here aims to facilitate both the validation and the correction of assemblies. The validation method was developed to locate putative assembly errors by gathering detailed quantitative information about each section of the assembly, and isolating the suspected misassemblies through the use of unsupervised learning algorithms. The data gathered during validation is then exploited in a second application designed for assembly correction. It is used to guide the process of meta-assembly, wherein two different assemblies are merged to produce a final sequence of greater quality. Supplementing the meta-assembly with detailed quality information allows for a more informed combination of sequences from each constituent assembly, resulting in a more accurate final product. Together, these applications aim to highlight the importance of incorporating sequence quality into assembly. The detailed detection of anomalous sequences, when combined with the capacity to also improve the assemblies, can enable the construction of more robust and correct assemblies.