To quantify the transcript level expression, raw reads were aligned using salmon [v1.9.0][cite:@patro2017] to the Bos taurus (ARS-UCD1.3,[cite:@bovine2009]) RefSeq transcripts with the genome used as decoys. The transcript expression was converted to gene-level expression using the R package tximport [v1.26.1] [cite:@soneson2016] using the edgeR scaling methods provided in the documentation. Genes with very low expression (fewer than 10 reads found in more than seven samples) were removed [cite:@bourgon2010]. The likelihood ratio method in edgeR [v3.40.2][cite:@mccarthy2012] was used to test for significant differential expression. The fry method from edgeR was used to do gene set enrichment analysis (GSEA) for Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO) terms. All code used for DE and GSEA can be found in XXX.
The 17 samples were sequenced to an average depth of 31.4 M reads (ranging from 29.4M in WT_2273 to 40.4M in WT_2226). Salmon mapped the raw reads to the bovine transcriptome with an average rate of 87.62% (81.12% in SLICK_2171 to 90.13% in SLICK_2287). The 76,278 reference transcript-level expressions for each sample were reduced to 20,946 gene-level expressions. Of these, 13,392 passed the low expression filter. A single gene (GeneID:513329, major allergen Equ c 1) showed significant (B.H. adjusted p-value <= 0.05) differential expression. Both the KEGG pathway and GO term GSEA showed no statistically significance enrichment at B.H. adjusted p-value <= 0.05.