The comparison of high-throughput single-cell RNA-seq methods

Here we compare the performance of these three approaches (inDrop, Drop-seq and 10x) using the same kind of sample with a unified data processing pipeline. We generated 2-3 replicates for each method using lymphoblastoid cell line GM12891. The average sequencing depth was around 50-60k reads per cell barcode. We also developed a versatile and rapid data processing workflow and applied it for all datasets. Cell capture efficiency, effective read ratio, barcode detection error and transcript detection sensitivity were analyzed as well. We used a human lymphoblastoid cell line GM12891 assuming homogeneous within the cell population throughout the experiments. Biological replicates were setup for all three methods, inDrop, Drop-seq and 10X Genomics Chromium (10X), with various cell inputs in different days and batches