Pattern discovery from biological data is crucial to advance our understanding of complex biological systems or biological processes and facilitate the application of our knowledge to benefit human needs. Base composition is an essential genomic feature. Findings of genome-wide base composition evolutionary pattern and its potential mechanisms can improve our understanding of genome evolution. Unmanned aerial vehicle-based high-throughput phenotyping platforms (UAV-HTPPs) can perform large-scale proximal measurements of phenotypic traits with high efficiency, high accuracy, and low cost, which provides novel opportunities to study the dynamic change of phenotypic traits across the growing season. The focus of my research is to study the genome-wide nucleotide evolutionary pattern following domestication in maize and soybean and time series normalized difference vegetation index (NDVI) data from a UAV-HTPP in maize.

We investigated the genome-wide base composition patterns through analyzing millions of SNPs segregating among 100 teosinte-maize accessions and among 302 wild-domesticated soybean accessions. Domesticated accessions have more nucleotide A and T across genome-wide polymorphic sites than wild accessions in maize and soybean. We demonstrated that different parts of the genome have differential contributions to the [AT]-increase between wild and domesticated accessions. The contribution to the [AT]-increase of non-genic part of the genome is greater than that of genic SNPs. The separation in [AT] values between wild and domesticated accessions is significantly enlarged in non-genic and pericentromeric regions. With motif frequency and sequence context analyses, we also showed that motifs (PyCG) related to solar-ultraviolet (UV) signature are enriched in non-genic and pericentromeric regions, particularly when they are methylated. Further genome scans using base-composition across polymorphic sites as a genome phenotype identify a set of putative candidate genes involved in UV damage repair pathways. Our findings establish important connections among UV radiation, mutation, DNA repair, methylation, and genome evolution.

Time series NDVI from 5 critical growth stages of 1,752 diverse maize accessions were extracted from spectral images acquired with a UAV-HTPP. We analyzed the dynamic change of NDVI across the growing season. Genotypic differences were identified with clustering analysis of time series NDVI. We conducted genome-wide association studies (GWAS) using static NDVI values from individual time points and growth curve parameters of NDVI dynamics across the growing season. GWAS with both static NDVI values and growth curve parameters identified a number of association signals. Additionally, GWAS with model fitted NDVI values discovered the dynamic change of the SNP effect for trait-associated genetic loci, which likely suggests the role of gene-environment interplay in affecting the development of NDVI across the growing season. Our results indicate that UAV-based remote sensing can assist the genetic dissection of NDVI.