Fig. 6

Identification and mechanisms of CBS variants that alter mRNA abundance. In all panels, data from the amplicon method is shown, and a gray dotted line indicates no change in variant effect or comparative metrics. For panel A, variants at a FDR < 0.1 are shown; in all other panels, variants with a FDR < 0.15 were analyzed. A CBS variant differential abundance. Structural residues near important features are labeled with an icon. Red and blue dotted lines represent the median for each input source. B Codon stability coefficient for variants grouped by directional effect. The difference in HEK293T ORFome codon stability coefficient (CSC) scores [52] between alternate and reference codons is compared. p-value indicates a one-sided Wilcoxon rank-sum test between RNA down and RNA up groups. C Comparison of codon stability between significant and non-significant variants. Significant variants were compared to another variant leading to the same amino acid change. Variants in the upper quartile of a null distribution are shown (“Methods”). D tRNA abundance correlation with mRNA abundance effects. The log fold change (logFC) is compared to the log ratio of tRNA abundance measured by Hydro-tRNAseq [56]. Spearman rank correlation (rho) and p-value are shown. Blue line is a fitted polynomial spline with a knot at 0. E Comparison of mRNA abundance effect with yeast functional complementation. Score (range 0–1, 0 is deleterious) is the max score of low and high vitamin B6 conditions [10]. Red text indicates a fitness score < 0.7 or a logFC < 0. Ref. = reference codon; Alt. = alternate codon; N.S. = not significant