Correlating pHi data with mRNA expression

Average pHi, standard deviation, and number of cells was determined for six different pHe values (from 6.63 – 7.70). Linear regression analysis based on pHi at all 6 different pHe values was carried out in MATLAB (Supplementary Code 3). Slope, intercept, resting pHi (at pHe 7.4), and standard deviation were obtained as an average for multiple biological repeats. Slope was plotted against standard deviation and k-means clustering was used to classify cell lines into distinct groups. For detection of differentially expressed genes in cell lines with high or low slope and standard deviations, t-tests were carried out (Supplementary Code 4).