Improving blood phenomics: ISU station report

Authors: TUGGLE, CHRISTOPHER - Iowa State University; CORBETT, RYAN - Iowa State University; YANG, PENGXIN - Iowa State University; DAHARSH, LANCE - Iowa State University; KAPOOR, MUSKAN - Iowa State University; HERRERA-URIBE, JUBER - Iowa State University; Byrne, Kristen; Loving, Crystal; LIM, KYU-SANG - Iowa State University; KOLTES, JAMES - Iowa State University; PAN, ZHANGYUAN - University Of California; WANG, YING - University Of California; GUAN, DAILU - University Of California; ESTRADA REYES, ZAIRA - University Of California; PROWSE-WILKINS, CLAIRE - University Of California; BI, YE - University Of California; AN, LIQI - University Of California; BAI, ZUECHEN - University Of California; ZHOU, HUAIJUN - University Of California; Nonneman, Danny - Dan; Smith, Timothy - Tim

Submitted to: Plant and Animal Genome Conference
Publication Type: Abstract Only
Publication Acceptance Date: 1/18/2023
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Molecular phenotypes (e.g. eQTL) are of interest in animal production to improve lowly heritable and difficult-to-measure traits, for precision management, and for understanding the impact of stress and immunity on important traits. Blood is a practical and available sample from which to collect such phenotypes, yet blood is exceedingly complex and heterogeneous, resulting in gene expression that is affected by both cell-type-specific gene expression and by cell composition. We have published scRNAseq of peripheral blood mononuclear cells as well as bulk RNAseq of nine cell types; however, these were collected from healthy pigs with no immune stimulation. To broaden the descriptive power of our dataset to recognize different cell type expression patterns, we are generating transcriptomes of all blood cell types across multiple conditions such as pre- and post-weaning and during pathogen challenges. We will then identify gene sets whose expression pattern across all cell types can be used to deconvolute blood samples (i.e., estimate cellular composition). Predictive gene lists will be independently validated and used to deconvolute a large (n= >1800 samples) blood RNA phenomics dataset on disease resilience traits. Such deconvolution will allow identification of cell-type specific gene expression and eQTL, and inform mechanistic interpretation of current and future eQTL and GWAS studies. Finally, we will develop methods to simplify future analyses of blood RNA phenomics by downsizing blood volume and gene set analyses. Overall, this research will improve the applicability and accuracy of whole blood RNA expression data to strengthen pig phenomics research.