Performance comparison across real and simulated datasets with different sequencing depths

A. Schematic representation of simulating scTCR-seq and scRNA-seq datasets. Red, green, orange, dark blue, light blue, black, and gray boxes represent V gene, CDR3 sequence, J gene, C gene, polyA sequence, cell barcode sequence, and non-TCR genes, respectively. B. Box plot showing the total read number of cells with successfully (True) and unsuccessfully (False) assembled TCRs on the collected experimental scRNA-seq datasets. The line in the middle represents the median; the bottom and the top of the box represent the 25th and 75th percentiles, respectively; and whiskers represent the minimum and maximum points within 1.5 times the interquartile range. C. Density plot showing the sequencing depth distribution of TCR sequences in TCR-assembled cells obtained from TRUST4. D. Bar plot showing the accuracy of paired CDR3 reconstruction on the simulated datasets with different sequencing depths. E. Bar plot showing the accuracy of paired J gene reconstruction on the simulated datasets with different sequencing depths. F. Bar plot showing the accuracy of paired V gene reconstruction on the simulated datasets with different sequencing depths. G. Bar plot showing the accuracy of paired AsTCR reconstruction on the simulated datasets with different sequencing depths. PolyA, a stretch of adenine-rich nucleotides added to the 3′ end of RNA molecules.