Fig. 5

The influence of the count thresholds on diversity matrices of TCR repertoire in sipuleucel-T treated prostate cancer patients (NeoACT study). a TCR sequencing data of PBMC samples at week 0 (PBMC.0), week 2 (PBMC.2) and week 4 (PBMC.4) of the five treated prostate cancer subjects are used for illustration. From top to bottom, each row shows the number of unique clones (Uniques), read depth, the Shannon index, Gini Simpson, Inverse Simpson (InvSimpson), geometric coefficient of variation (GCV) and Clonality of TCR repertoire. From the left to the right, each column presents the different threshold of the clonotypes count (original data which is > =2, > = 5, > = 10, > = 15, > = 20, > = 25 and > =30). The Shannon index, Clonality, Gini Simpson, Inverse Simpson and GCV were obtained by recalculating the clone frequency after filtering the data with the different cutoffs. b Pairwise relative clonality were calculated as the clonality of PBMC at the later time point divided by that of the earlier time point, e.g., PBMC.2/0 = clonality of PBMC Week 2 divided by PBMC Week 0. From the left to the right, each column presents the different threshold of the clonotypes count (original data which is > =2, > = 5, > = 10, > = 15,> = 20, > = 25 and > =30). The subject with triangle shapes was the example used in Fig. 1c)