Computational protocol: Massively parallel whole genome amplification for single-cell sequencing using droplet microfluidics

Similar protocols

Protocol publication

[…] In the 1st droplet generator, cell suspension and buffer D2 reagent (QIAGEN, Hilden, Germany) for cell lysis were pumped from different two inlets into the cross-junction as dispersed-phase liquids, while the carrier phase fluorinated oil HFE7500 containing 2% (v/v) of the surfactant Pico-Surf1 (Dolomite, Charleston, MA, USA) was driven from the other inlet using Mitos P-Pump (Dolomite). The cell suspension and lysis reagent met at the cross-junction, and droplets were periodically pinched off from the dispersed phase, at flow rates of 24 μL/h for both dispersed-phase liquids and 360 μL/h for the carrier oil. The device outlet was also connected to a collecting PCR tube via PTFE tubing. The total 10–40 μL of cell aqueous solutions were converted into approximately 0.3–1.2 × 106 droplets (diameter: 40 μm). The cells encapsulated into individual droplets were then incubated at 65 °C in PCR tubes for 10 min. [...] In order to confirm amplification from single bacteria, 16S rRNA gene identification was also performed against SAGs obtained after the 2nd-round MDA. Primer pair sequences for the V3 and V4 region were used according to the protocol of Illumina Miseq system for 16S metagenomics sequencing library preparation. The PCR amplicons of 16S rRNA gene fragments (V3-V4 region) were sequenced by MiSeq (Illumina, San Diego, CA, USA) according to the Illumina protocol. For 16S analysis of soil metagenome, RDP classifier 2.11 was used for taxonomic assignments. [...] For bacteria analysis, acquired reads were normalized to 1×, 5×, 10×, 20×, 40×, and 60× coverage per each genome for each sample. All sequence data were mapped to the NCBI reference genome of NC_00913 (E. coli substrain MG1655) or NCBI reference genome of NC_014479 (Bacillus subtilis subsp. spizizenii str. W23) using the software BWA. Genome coverage was calculated using SAMtools. Each normalized read was assembled de novo using SPAdes 3.5.0, and the contigs were qualified by QUAST 2.3. CheckM 1.0.6 was used to assess the completeness and the contamination read rate after contigs smaller than 2,000 bp were removed. For SAG assembly, reads from single E. coli cells or B. subtilis were subsampled to 60× coverage per genome. These subsampled reads were combined together and assembled with SPAdes.For cancer cell analysis, acquired reads were mapped to the UCSC human genome 19 (hg19) using BWA. After removal of reads with mapping quality score less than 40 from created SAM files, genome coverage was calculated using SAMtools. The Genome Analysis Toolkit (GATK) was used to detect single nucleotide variants (SNVs) of single cancer cells and population data (NCBI SRA053195). The GATK UnifiedGenotyper was used to generate a single Variant Call Format file from all single cells and population data. Detected variants were recalibrated using the GATK VariantRecalibrator and database constructed from hapmap 3.3, dbSNP build 138, Omni 2.5 M chip, and Mills. Using recalibrated SNVs, allelic dropout rate (ADR) and false positive rate (FPR) were calculated. […]

Pipeline specifications

Software tools MITOS, RDP Classifier
Applications Genome annotation, Metagenomic sequencing analysis, 16S rRNA-seq analysis
Organisms Bacteria, Homo sapiens
Diseases Neoplasms