Fig. 2

Evidence for excessive copy number variation in the C. felis genome. a Flea genome size estimates. Flow cytometer-based estimates were performed for male and female individuals of X. cheopis (Texas) and C. felis (Texas), and for female C. felis EL from the sequenced colony (see Additional file 3: Fig. S2). The inset (top right) depicts the coefficients of variation in measured fluorescence (relative fluorescence units (RFU)) for Drosophila melanogaster (n = 26), D. viridis (n = 26), and C. felis EL (n = 26) females prepared and analyzed simultaneously. b Graphic depiction of assembling CNV. Two theoretical individual fleas are shown with different CNVs for loci A and B. Regions unique to each individual genome are shown by the red dashed boxes. Only reads concordant between individuals are included in the conglomerate assembly. c Comparison of Illumina read coverage mapping between duplicate genes (blue) and single-copy genes (green) at different %ID thresholds. Reads that mapped to multiple locations (alternative mappings) were included. Asterisks indicate a statistically significant difference (Welch two-sample t test, p < 2.2e−16) between mean coverage of single-copy and duplicate genes at the 90%ID threshold. d Transcriptional support for C. felis EL genes within the 1KITE transcriptomic data. Counts of transcripts per million reads (TPM) were mapped (Hisat2 and Stringtie), binned, and plotted against the number of duplicated (blue) and single-copy (green) genes in the BIG9 assembly. e The extent of truncation within clusters of duplicated genes in C. felis. The number of clusters with truncated members at each integer %ID threshold (left) was calculated as the proportion of the total clusters at that threshold (center). The distribution of length differences in these clusters (relative to the longest member in each cluster) is plotted as a violin plot (right); black diamonds represent the mean length difference at each %ID threshold