Fig. 1

Exhaustive capture of MAPs RNA expression. a Overview of the BamQuery approach to measuring MAP RNA expression levels. b Pearson’s correlation between BamQuery-acquired read counts and Jellyfish’s K-mer counts for canonical nonamer MAPs (n = 1211) from the HLA Ligand Atlas (present in at least 20 different tissues) in eight mTEC samples. c Pearson’s correlation between BamQuery’s (in RPHM) and Kallisto’s (in TPM) quantifications of 1702 MAPs from the HLA Ligand Atlas in 8 mTEC samples. Because Kallisto does not perform direct quantifications of MAPs’ RNA expression, the expression of their gene of origin was used as a surrogate. A value of 0.5 was added to each RPHM or TPM value to enable visualization on a logarithmic axis. Correlations for three representative samples and the average of the eight samples are shown. d Segregation of MAPs based on their number of coding regions and re-computation of correlations for the average of eight mTEC samples. e Box plots of the average expression of MAPs across the eight mTECs, segregated based on the number of coding regions. f Correlation between the ratio, for each MAP, between the BamQuery and the Kallisto quantification (average of eight mTECs), as a function of the number of coding regions. g Correlation between BamQuery’s (in RPHM) and Kallisto’s (in TPM) quantifications of 108 mutated MAPs in eight mTEC samples (analysis performed as in panel c). h The number of mutated MAPs having an expression = 0 according to Kallisto or BamQuery is reported in each of the eight tested mTEC samples (dots)