Saccharomyces beer yeasts are well-known for their sensory contributions to beer, but the molecular and genetic factors underpinning differences in fermentation performance and flavor profiles between yeast strains are not fully understood. Genomic comparisons offer a comprehensive, albeit static picture of the similarities and differences between ale and lager strains. In contrast, transcriptome analysis captures genetic expression in growth-specific and environmental contexts, including those relevant to fermentation such as log-phase growth and senescence. In this study, we employ comparative transcriptomics and interpretable machine learning methods to classify and highlight differences in beer-relevant gene transcription between eight brewing yeast strains with robustness to different fermentation temperatures and growth media. Our computational pipeline, herein referred to as YeasTSP, reveals discriminative patterns in expression of genes important in beer production, such as alcohol production, acetate ester biosynthesis, and thiol production. Although our study is based on a small case study sample set, the approach focuses specifically on genes relevant to yeast performance and potentially enables downstream applications like monitoring yeast performance, relating wild and engineered strains to existing strains, and identifying latent similarities and differences between strains.