Naina Kumari

Livestock are important drivers towards sustainable development goals through promoting resilience, productivity in small farmers, and involvement in markets. Cattle (Bos taurus) and buffalo (Bubalus bubalis) have important roles to play in Asian and Indian economy through other significant products in addition to milk and meat. Due to low availability of genomics and transcriptomics resources on buffaloes, genetic improvement efforts are hindered. To fill this gap, we compared 2,429 transcriptomes from 438 BioSamples in 23 BioProjects, spanning 76 river and swamp buffalo tissues and cell types. This prompted the creation of BuffExDB (http://46.202.167.198/buffex/), an easy-to-use, filterable database with tissue-specific gene expression, provides Tau scores for tissue-specific genes including functional annotations. This is the first of its kind to provide an easily browsable and filterable database that allows users to view and visualize the expression level of each tissue in multiple samples at once. In addition, we have performed meta-analysis of bovine transcriptome datasets to determine crucial genes involved in bovine tuberculosis (BTB) in cattle by combining multiple independent studies using a unified bioinformatics workflow. In the present research, we determined major genes, pathways, and ontologies in relation to BTB disease process. RNA-Seq technology has revolutionized transcriptomic research with insights into gene expression in varying biological conditions However, optimizing RNA-Seq experimental design remains a challenge, particularly in determining appropriate sample sizes based on heritability and statistical power. We created a statistical tool based on a linear model to calculate RNA-Seq sample sizes from heritability, the first of its kind. This method considers false discovery rates, heritability, tissue type, fold change, power and sample-to-sample variation, making differential expression studies more reliable. The findings of the study reveal that sample size is inversely proportional to trait heritability i.e. when the heritability is low, a higher number of replicates should be used in order to achieve the required statistical power as compared to medium and high heritable traits. To further assist researchers, we introduced HEssRNA (https://cran.r-project.org/web/packages/HEssRNA/index.html) and HEssRNA-Shiny, an R package created in CRAN repository and a web-based shiny tool respectively for sample size estimation in RNA-Seq studies based on bovine gene expression data using the model developed. The web tools are an easy-to-use resource for non-programmers to estimate sample size based on heritability. Both the package and tool offer option for power calculation starting from RNA-Seq count matrix. Although designed for bovine data, the tools can be customized for other species based on input data and heritability values. Collectively, these resources form a strong platform for transcriptomic studies, enabling data-informed experimental design and enhancing the reproducibility of gene expression research in cattle and buffalo. Our research helps to advance bovine functional genomics and enables precision livestock research.