作者：ahworld
链接：vartrix | Single-Cell Genotyping Tool
来源：anno-sc.com
著作权归作者所有，任何形式的转载都请联系作者。
VarTrix是一款10XGenomics官方的软件用于从10XGenomics单细胞数据中提取单细胞变异信息。
- VarTrix 并不执行
variant calling - 需要输入先前定义的
variant calls - VarTrix可用于评估样品中单细胞的异质性
VarTrix 流程图
image
github
https://github.com/10XGenomics/vartrix
下载安装
VarTrix使用rust语言编写，官方提供了编译后的二进制文件，直接下载就能使用。
image
我们以linux版为例，下载后增加可执行权限。
chmod +x vartrix_linux
测试
下载测试数据
为了快速理解VarTrix的使用，我们从github下载其测试数据
https://github.com/10XGenomics/vartrix/test
test目录里提供了多套测试数据，我们选下面这一组输入文件做测试：
- test_dna.vcf
单样品的VCF格式变异文件，用于输入 - test_dna.bam，test_dna.bam.bai
Input Cell Ranger BAM，需要有index文件 - test_dna.fa, test_dna.fa.fai
FASTA格式的参考基因组文件，需要有index文件 - dna_barcodes.tsv
Cell Ranger输出的矩阵barcode文件
使用说明
./vartrix_linux -h查看程序使用说明
vartrix 1.1.14
Ian Fiddes <ian.fiddes@10xgenomics.com> and Patrick Marks <patrick@10xgenomics.com>
Variant assignment for single cell genomics
USAGE:
vartrix_linux [FLAGS] [OPTIONS] --bam <FILE> --cell-barcodes <FILE> --fasta <FILE> --vcf <FILE>
FLAGS:
-h, --help Prints help information
--no-duplicates Do not consider duplicate alignments
--primary-alignments Use primary alignments only
--umi Consider UMI information when populating coverage matrices?
-V, --version Prints version information
OPTIONS:
-b, --bam <FILE> Cellranger BAM file
--bam-tag <bam_tag> BAM tag to consider for marking cells? [default: CB]
-c, --cell-barcodes <FILE> File with cell barcodes to be evaluated
-f, --fasta <FILE> Genome fasta file
--log-level <log_level> Logging level [default: error] [possible values: info, debug, error]
--mapq <INTEGER> Minimum read mapping quality to consider [default: 0]
-o, --out-matrix <OUTPUT_FILE> Output Matrix Market file (.mtx) [default: out_matrix.mtx]
--out-variants <OUTPUT_FILE> Output variant file. Reports ordered list of variants to help with loading into
downstream tools
-p, --padding <INTEGER> Number of padding to use on both sides of the variant. Should be at least 1/2 of read
length [default: 100]
--ref-matrix <OUTPUT_FILE> Location to write reference Matrix Market file. Only used if --scoring-method is
coverage [default: ref_matrix.mtx]
-s, --scoring-method <scoring_method> Type of matrix to produce. In 'consensus' mode, cells with both ref and alt reads are
given a 3, alt only reads a 2, and ref only reads a 1. Suitable for clustering. In
'coverage' mode, it is required that you set --ref-matrix to store the second matrix
in. The 'alt_frac' mode will report the fraction of alt reads, which is effectively the
ratio of the alternate matrix to the sum of the alternate and coverage matrices.
[default: consensus] [possible values: consensus, coverage, alt_frac]
--threads <INTEGER> Number of parallel threads to use [default: 1]
--valid-chars <valid_chars> Valid characters in an alternative haplotype. This prevents non sequence-resolved
variants from being genotyped. [default: ATGCatgc]
-v, --vcf <FILE> Called variant file (VCF)
测试代码
./vartrix_linux -v ./test_dna.vcf -b ./test_dna.bam -f ./test_dna.fa -c ./dna_barcodes.tsv -o ./matrix.mtx
vartrix的输出文件格式与Cell Ranger类似，
- 第1列：VCF中的SNP所在行号
- 第2列：barcode在barcodes.tsv中的行号
- 第3列：相应barcode（细胞）SNP鉴定结果
其中第3列的值只有0，1，2，3这四种，分别代表：
- 0 : No Call
- 1 : ref/ref
- 2 : alt/alt
- 3 : alt/ref
# head(./matrix.mtx)
%%MatrixMarket matrix coordinate real general
% written by sprs
46 1331 1073
1 8 1
1 73 1
1 154 2
1 174 1
1 191 2
1 194 2
1 243 1
UMAP图中展示VarTrix结果
要想在各细胞亚群UMAP图中展示VarTrix结果，我们需要先用seurat对细胞做常规分群分析，然后把VarTrix结果作为MetaData添加给Seurat对象。
VarTrix结果转为可添加给Seurat对象的MetaData
variants文件转为类似于geneID的格式，相当于给每个变异一个唯一的编号（以染色体及变异具体位置命名）
awk '{print $1,$2}' test_dna.vcf > SNV.loci.txt
sed -i 's/\s/:/g' SNV.loci.txt
head SNV.loci.txt
#CHROM:POS
1:13116
1:13118
1:13656
1:15211
1:15274
读取VarTrix的结果matrix, barcodes和variants
library(Seurat)
library(Matrix)
library(stringr)
setwd("~/Documents/RD/Rs/VarTrix")
snv_matrix <- readMM("matrix.mtx")
# convert the matrix to a dataframe
snv_matrix <- as.data.frame(as.matrix(t(snv_matrix)))
#read in the cell barcodes output by Cell Ranger
barcodes <- read.table("dna_barcodes.tsv", header = F)
# read in SNV loci
# Should be constructed a single column. For example
# chr1:1234-1235
# chr2:2345-2346
# Construct the final table to add to the Seurat object
snps <- read.table("SNV.loci.txt", header = F)
colnames(snv_matrix) <- barcodes$V1
row.names(snv_matrix) <- snps$V1
snv_matrix_t <- as.data.frame(t(snv_matrix))
snv_matrix_t <- str_replace(as.character(snv_matrix_t), "0", "No Call")
snv_matrix_t <- str_replace(as.character(snv_matrix_t), "1", "ref/ref")
snv_matrix_t <- str_replace(as.character(snv_matrix_t), "2", "alt/alt")
snv_matrix_t <- str_replace(as.character(snv_matrix_t), "3", "alt/ref")
添加VarTrix信息给Seurat对象
Seurat.obj<-readRDS("/path/Step5_combinedObj.rds")
Seurat.obj <- AddMetaData(object = Seurat.obj, metadata = snv_matrix_t)
UMAP展示
UMAPPlot(object = Seurat.obj, group.by = "1:13116")
UMAP展示结果
因为VarTrix的github并未提供基因表达矩阵，所以VarTrix结果在UMAP中的展示我们暂用模拟数据代替
下面是模拟VarTrix结果添加给Seurat对象及可视化代码。
Seurat.obj<-readRDS("~/Desktop/blish_covid.seu.rds")
gts <- data.frame(snp_0001=rep(0, times=length(names(Seurat.obj$nFeature_RNA))),snp_0002=rep(1, times=length(names(Seurat.obj$nFeature_RNA))))
rownames(gts) <- names(Seurat.obj$nFeature_RNA)
gts[sample(rownames(gts),4000), "snp_0001"] <- 1
gts[sample(rownames(gts),5000), "snp_0001"] <- 2
gts[sample(rownames(gts),6000), "snp_0001"] <- 3
gts$snp_0001 <- str_replace(as.character(gts$snp_0001), "0", "No Call")
gts$snp_0001 <- str_replace(as.character(gts$snp_0001), "1", "ref/ref")
gts$snp_0001 <- str_replace(as.character(gts$snp_0001), "2", "alt/alt")
gts$snp_0001 <- str_replace(as.character(gts$snp_0001), "3", "alt/ref")
Seurat.obj <- AddMetaData(object = Seurat.obj, metadata = gts)
UMAPPlot(object = Seurat.obj, group.by = "snp_0001")
image
其中blish_covid.seu.rds来自下面这篇文章
Wilk A J, Rustagi A, Zhao N Q, et al. A single-cell atlas of the peripheral immune response in patients with severe COVID-19[J]. Nature Medicine, 2020: 1-7.
2020是不平凡的一年，在年终来临之际，我们计划组织一次单细胞线上高级分析培训班，希望能够给您提供一点参考，我们将以上面这篇文章的数据为基础线索进行高级分析的代码重现，另外还有marker gene可视化应用代码的赠送。
