The genetic improvement of Manihot esculenta, or cassava, has historically been slow, largely because its biology renders traditional breeding techniques inefficient and because of little interest from the private sector. The goal of the Next Generation Cassava Breeding project (NEXTGEN) is to assist breeding institutions in Nigeria, Uganda, and Tanzania with increasing the rate of genetic improvement of cassava through implementation of genomic selection (GS). The three chapters of my thesis outline my work and involvement with the NEXTGEN project. The first chapter details our investigation of two questions: 1) can we use existing imputation methods developed by the human genetics community to impute missing genotypes in datasets derived from non-human species and 2) are these methods, which were developed and optimized to impute ascertained variants, amenable for imputation of missing genotypes at next-generation sequencing (NGS)-derived variants? In the second chapter, we introduce a statistical method, BIGRED (Bayes Inferred Genotype Replicate Error Detector), for detecting mislabeled and contaminated samples using shallow-depth sequence data. BIGRED addresses key limitations of existing approaches and produced highly accurate results in simulation experiments. In the third chapter, we outline how we used the multi-generational pedigree and genotyping-by-sequencing (GBS) data from the International Institute of Tropical Agriculture (IITA) to characterize the recombination landscape across the 18 chromosomes of cassava. We detected SNP intervals containing crossover events using SHAPEIT2 and duoHMM, constructed a genetic map using these intervals, compared it to an existing map constructed by the International Cassava Genetic Map Consortium (ICGMC), and constructed sex-specific genetic maps to see if cassava displays sexual dimorphism in crossover distribution and frequency.