Fig. 4

Analysis of binding neoantigens to patient HLA alleles. a Empirical cumulative distribution function showing the proportion of expressed missense and indel mutations labeled as binding neoantigens at different PHBR-I score cutoffs in MSS patients. b Boxplots comparing the fraction of binding neoantigens in tumors with no B2M or HLA mutation (teal) versus patients with a B2M mutation (yellow) or HLA mutation (purple). A PHBR-I score cutoff of 2 was used to designate a binding neoantigen for this comparison. c Total number of neoantigens that bind to a patient’s mutated HLA allele versus the average number of neoantigens across the unmutated HLA alleles across all cancer types for MSS patients. A red line indicates that there are more neoantigens with binding affinity to the mutated HLA allele than the average across the unmutated HLA alleles; and a blue line depicts the opposite trend. d Allelic fraction percentile distribution for expressed mutations in MSS patients with B2M and HLA mutations. We used the Kolmogorov-Smirnov statistic to determine whether the two distributions were significantly different. e Comparison of the number of expressed neoantigens with binding affinity to the patient-specific mutated allele between the low AF percentile (< 40%) and the high AF percentile (> 60%) HLA mutated patients. Patients with MSI and with mutations in both B2M and HLA genes were excluded. f Comparison of the total number of neoantigens that bind to a patient-specific mutated HLA allele versus the average number of neoantigens with binding affinity to the five unmutated HLA alleles in patients with high allelic fraction percentile HLA mutations