SCCNV: a software tool for identifying copy number variation from single-cell whole-genome sequencing

Version 1.0.2

Updated date: 2020.07.20

Author: Xiao Dong

Email: [email protected], [email protected]

Licensed under the GNU Affero General Public License version 3 or later

python 2 or python 3

python modules os, argparse, math, numpy

bedtools, samtools

-i, (or --ibam): a text file providing a list of bam files of single-cell whole-genome sequencing data; one row one file; required input file.

-o, (or --odir): a directory to output result; required output directory

-g, (or --genome): genome version; default: hg19

-s, (or --genomesize): genome size; default: 3140000000

-b, (or --binsize): binsize (bp); default: 500000; min: 1000

-w, (or --windowsize): windowsize (number of bins); default: 10

-p, (or --ploidy): ploidy of autosomes; default: 2

-m, (or --minmap): minimum mappability; default: 0.3

-n, (or --maxN): maximum N in the bin; default: 0.1

-k, (or --skipbed): skip calulcating bed with bedtools; default: False

-r, (or --report): report intermediate results; default: False

Below is an example for cell_A. I recommend filter out reads with mapq<30.

samtools view -b -q 30 ./bam/cell_A.bam > ./bam_mapq30/cell_A.mapq30.bam

samtools index ./bam_mapq30/cell_A.mapq30.bam

Provide a file (e.g. “bamlist.txt”) with the following content. See NOTE below for important information when analyzing multiple single cells of a same clone, e.g., tumor cells.

./bam_mapq30/cell_A.mapq30.bam

./bam_mapq30/cell_B.mapq30.bam

./bam_mapq30/cell_C.mapq30.bam

python sccnv.py -i bamlist.txt -o cellsAtoC -k False -r True

One can prepare bed files for calculation using bedtools and samtools (for every cell in paralleles),

mkdir cellsAtoC

bedtools makewindows -g ./resource/hg19.chrlength.txt -w 500000 > ./cellsAtoC/hg19.bin500000.bed

samtools bedcov ./cellsAtoC/hg19.bin500000.bed ../bam_mapq30/cell_A.mapq30.bam > ./cellsAtoC/cell_A.depth.bin500000.bed

And skip the related step in SCCNV,

python sccnv.py -i bamlist.txt -o cellsAtoC -k True -r True

An example dataset is included in "sccnv_example_v1.0.2.zip". It includes (1) intermediate and final result files generated using SCCNV in txt format; and (2) basic R scripts for visualizing CNVs across the genome.

(1) SCCNV aims to discover difference in CNV between every single cell in the bamlist.txt and the other cells in the bamlist.txt. When analyzing CNV of multiple tumor cells, it is not appropriate to include all tumor cells in the bamlist.txt. Instead, please use one tumor cell with two or more normal diploid cells in the bamlist.txt.

v1.0.2 2020.07.20, allowed calling CN>4 and included additional example data.

v1.0.1 2020.04.10, revised readme and included example data.

v1.0.0 2018.07.26, release version.

v0.0.3 2018.07.25, rewrote in Python.

v0.0.2 2017.08, fixed bugs.

v0.0.1 2016.09, drafted script in R.