Fig. 2

Results from the mutscan re-analysis of the FOS/JUN protein interaction data set from [6]. A The number of variants detected with different numbers of mutated bases. B The average abundance of variants with different numbers of mutated bases. While a larger number of different variants with two or more mutations are observed, these are generally much less abundant than variants with no or a single mutated base. C, D Diagnostic plots of read filtering performed by digestFastqs(). In this data set, the main reasons for filtering out read pairs are that they either contain an adapter sequence or that the number of mutated codons exceeds the defined threshold of maximum one mutated codon per protein. E Pairs plot displaying the concordance of the observed counts across the six samples. Generally, high correlations are observed between the three input samples as well as between the three output samples, indicating good robustness of the measurements. The correlation between the input and output samples is considerably lower, reflecting the impact of the selection. F Heatmap showing the estimated protein–protein interaction score for all single-amino acid variants of each of the two proteins. Red color indicates an increased interaction, while blue signifies decreased interaction