Sequencing data was processed using a customized bioinformatics pipeline (101-pipeline) designed to detect several classes of genomic variants, including single-nucleotide variants (SNV), small insertions and deletions (InDel), copy number variants (CNV), and fusions. The raw sequencing data were preprocessed by using bcl2fastq. Preprocessed sequencing data were mapped to the human genome (hg19) by using Burrows-Wheeler Aligner 0.7.10 to generate BAM files. Variant calling was performed by using Vardict to detect SNVs/InDels. The false positive SNVs/InDels identified in cfDNA from healthy donors were used for filtering of background variants. Variants were annotated with ANNOVAR and SnpEff v3.6. The maximum allele frequencies (AF) of somatic variants were calculated. Factera was used to identify genomic fusions and calculate their abundance. The threshold for fusion detection is pair count ? 1 and total count ? 3. Large genomic rearrangement (LGR) is defined as the deletion or duplication of single- or multi-exons at the gene level. 101-test indicated a NGS test on 101-panel to identify somatic variants only in ctDNA sample without paired whole blood cells (WBCs) sample using 101-pipeline. Conventional test indicated using digital droplet polymerase chain reaction (ddPCR)/break point PRC to identify somatic variants in ctDNA sample with paired WBCs sample.