Genomic Prediction (GP) is commonly performed using tens of thousands, even millions of single-nucleotide polymorphism (SNP) markers. Associations among the phenotypes and genotypes of a training population are used to predict the performance of un-phenotyped target populations. Traditionally, the markers used in these studies are treated similarly, irrespective of their position in the genome, their proximity to regulatory elements, or whether they reside within biologically-relevant genes. The content of this dissertation aims to both identify “functional” regions in the cassava (Manihot esculenta) genome and test whether the incorporation of prior biological knowledge enhances the accuracy of GP models in this crop. In 2013, at the onset of this research, very few genomic resources were available within the cassava research community; a draft of the genome sequence had recently been released and a solid platform for low-coverage genotyping using genotyping-bysequencing (GBS) was online. As a means of generating more genomic resources, we first began by identifying nucleotide-binding site leucine-rich repeat (NBS-LRR) genes associated with biotic resistance across the cassava genome. We then leveraged the NBS-LRR information, together with genomic annotations and a transcriptomics experiment, in a second study to identify genes involved in the interaction of cassava with Cassava Brown Streak Virus (CBSV). We later used biologically-informed GP methods to compare models with and without biologically-relevant information. Until the final phase of our research, our efforts had focused on identifying functional elements within the coding fraction of the genome. In an effort to build upon several genome-wide association (GWA) studies illustrating the importance of regulatory regions outside genes, the third study explored cassava’s nascent transcriptome. In doing so, we were able to identify key components of plant transcriptional regulation and candidate enhancer regions that not been previously described. Moreover, we showed that these candidate enhancer regions contributed disproportionately to the SNP heritability of several complex traits. The research presented herein provides holistic insight into cassava’s genomic resources, and it is our hope that it is useful to future research and breeding endeavors within this staple crop species.