单细胞转录组跨物种比对

一、写在前面

在跨物种单细胞研究中，如何消除物种间的进化差异并实现细胞类型的精确比对，是单细胞转录组（scRNA-seq）分析中的难点。针对南极鱼（Trematomus bernacchii, TB）与尼罗罗非鱼（Oreochromis niloticus, ON）的脑组织数据，本教程展示了从 Cell Ranger 原始流程启动，到利用Diamond进行跨物种同源基因预测，并最终通过CCA (Canonical Correlation Analysis) 算法实现多物种数据集整合的完整管线。

在分析过程中，我们不仅引入了Scrublet算法来精准拦截模拟双细胞，还针对非模式物种研究中常见的基因注释不全问题，演示了如何通过解析GFF3注释文件自定义基因名映射表。通过对比Harmony与CCA在跨物种整合上的表现，让我们在复杂进化背景下，从数万个细胞中抽丝剥茧，定位保守的神经元亚群及其差异表达特征。

如果需要单细胞数据分析 指导、生信热点全文复现 、自测数据个性化分析 辅导、常态化实验学习 ，欢迎联系[Biomamba_zhushou]。

二、实操流程

1.准备工作

1.1 安装cellranger

下载解压安装包

复制代码

cd /storage/public/yongpeng/work/cellranger

wget -O cellranger-10.0.0.tar.gz "https://cf.10xgenomics.com/releases/cell-exp/cellranger-10.0.0.tar.gz?Expires=1775781670&Key-Pair-Id=APKAI7S6A5RYOXBWRPDA&Signature=Z51eZif2whpR7Rmp-It1rTPG-P2Z9UOHm-MgqSjds~AJs3cuf98cL0a1oLI8neaeZom72Sr9PQQv8KR3bbWSz1qpNuCMbkgtlG0l7SUC64q7ZTAR8oBmMd~lyfY5961Ba9dnPPtr-JS94SCKHnaAtzv6LfDoWUIu2HZkTi8~eWufI8MjxZCWaMldoQCOkxUBtTwnTUKR-EZjSKCX2ZAt7Ip3-dWcVb2QayDesuRbP~MrN7akrqrCY8~Lfcuv2H7rmc3~yeuklrg0A~DekYqjXny--bouLBVnTzUhyTUx8wZgjgVPBralpU~RuoJkS6Yaklu1nSGpcTK71ozwjhl61A__"

tar -zxvf /storage/public/yongpeng/work/cellranger/cellranger-7.1.0.tar.gz

配置软件路径以及检查是否安装成功

配置软件路径方便后续直接使用cellranger

复制代码

echo 'export PATH=/storage/public/yongpeng/work/cellranger/cellranger-7.1.0:$PATH' >> ~/.bashrc

source .bashrc

尝试使用cellranger看是否安装成功

复制代码

cellranger

cellranger cellranger-7.1.0 Process 10x Genomics Gene Expression, Feature Barcode, and Immune Profiling data

USAGE: cellranger

OPTIONS: -h, --help Print help information -V, --version Print version information

SUBCOMMANDS: count Count gene expression (targeted or whole-transcriptome) and/or feature barcode reads from a single sample and GEM well multi Analyze multiplexed data or combined gene expression/immune profiling/feature barcode data multi-template Output a multi config CSV template vdj Assembles single-cell VDJ receptor sequences from 10x Immune Profiling libraries aggr Aggregate data from multiple Cell Ranger runs reanalyze Re-run secondary analysis (dimensionality reduction, clustering, etc) targeted-compare Analyze targeted enrichment performance by comparing a targeted sample to its cognate parent WTA sample (used as input for targeted gene expression) targeted-depth Estimate targeted read depth values (mean reads per cell) for a specified input parent WTA sample and a target panel CSV file mkvdjref Prepare a reference for use with CellRanger VDJ mkfastq Run Illumina demultiplexer on sample sheets that contain 10x-specific sample index sets testrun Execute the 'count' pipeline on a small test dataset mat2csv Convert a gene count matrix to CSV format mkref Prepare a reference for use with 10x analysis software. Requires a GTF and FASTA mkgtf Filter a GTF file by attribute prior to creating a 10x reference upload Upload analysis logs to 10x Genomics support sitecheck Collect linux system configuration information help Print this message or the help of the given subcommand(s)

表明安装成功

1.2 准备单细胞测序数据

复制代码

ls /storage/public/yongpeng/work/cellranger/ONRNA

T2RNA_S1_L001_R1_001.fastq.gz T2RNA_S2_L001_R2_001.fastq.gz T2RNA_S4_L001_R1_001.fastq.gz T2RNA_S1_L001_R2_001.fastq.gz T2RNA_S3_L001_R1_001.fastq.gz T2RNA_S4_L001_R2_001.fastq.gz T2RNA_S2_L001_R1_001.fastq.gz T2RNA_S3_L001_R2_001.fastq.gz

复制代码

ls /storage/public/yongpeng/work/cellranger/TBRNA

TBRNA_S1_L001_R1_001.fastq.gz TBRNA_S2_L001_R2_001.fastq.gz TBRNA_S4_L001_R1_001.fastq.gz TBRNA_S1_L001_R2_001.fastq.gz TBRNA_S3_L001_R1_001.fastq.gz TBRNA_S4_L001_R2_001.fastq.gz TBRNA_S2_L001_R1_001.fastq.gz TBRNA_S3_L001_R2_001.fastq.gz

1.3 下载基因组文件和注释文件

ON数据（Oreochromis_niloticus）

复制代码

wget https://ftp.ensembl.org/pub/release-115/fasta/oreochromis_niloticus/dna/Oreochromis_niloticus.O_niloticus_UMD_NMBU.dna.toplevel.fa.gz 
wget https://ftp.ensembl.org/pub/release-115/gtf/oreochromis_niloticus/Oreochromis_niloticus.O_niloticus_UMD_NMBU.115.gtf.gz

TB数据（Trematomus_bernacchii）此基因组文件和注释文件是实验室组装制作

1.4 运行cellranger

复制代码

nohup cellranger mkref \
    --genome=Oni_genome_rna \
    --fasta=Oreochromis_niloticus.O_niloticus_UMD_NMBU.dna.toplevel.fa \
    --genes=Oreochromis_niloticus.O_niloticus_UMD_NMBU.111.gtf &

nohup /cellranger count \
--id=ON_cellranger\
--transcriptome=/storage/public/yongpeng/work/cellranger/Oni_genome_rna \
--fastqs=/storage/public/yongpeng/work/cellranger/ONRNA \
--sample=T2RNA \
--create-bam=true > $s.log 2>&1 &

nohup cellranger mkref \
    --genome=TB_genome_rna \
    --fasta=Trematomus_bernacchii.dna.fa \
    --genes=Trematomus_bernacchii.gtf &

nohup cellranger count \
--id=TB_cellranger \
--transcriptome=/storage/public/yongpeng/work/cellranger/TB_genome_rna \
--fastqs=/storage/public/yongpeng/work/cellranger/T2RNA \
--sample=TBRNA \
--create-bam=true > $s.log 2>&1 &

ls /storage/public/yongpeng/work/cellranger/ON_cellranger

cmdline _finalstate _log _perf _sitecheck _timestamp _versions extras _invocation _mrosource _perf._truncated _tags _uuid _filelist _jobmode outs SC_RNA_COUNTER_CS ON.mri.tgz _vdrkill 检查目录是否完整

若要合并不同物种的数据，需要先统一基因名，这里通过寻找同源基因来达成目的，由于ON物种基因研究较为全面因此把TB基因转换成ON对应的同源基因。

1.5 注释同源基因

准备两个物种的蛋白序列文件TB是实验室提供，ON需要从ENSEMBL下载

复制代码

cd /home/yongpeng/work/blast/diamond

wget https://ftp.ensembl.org/pub/release-115/fasta/oreochromis_niloticus/pep/Oreochromis_niloticus.O_niloticus_UMD_NMBU.pep.all.fa.gz

Eggnog的比对结果相比于diamond更准确，但是由于运行时间过长，这里仅使用diamond作为示例

复制代码

nohup diamond makedb --in Oreochromis_niloticus.O_niloticus_UMD_NMBU.pep.all.fa.gz --db ON_diamond &

nohup diamond blastp -q Trematomus_bernacchii.pep.faa --db ON_diamond --out /home/yongpeng/work/blast/diamond/TB_ON -p 4 &

提取出基因对应表（每个对应关系只保留匹配度最高，匹配度低于80%则删除）保存为TB_TO_ON.xlsx

2.单细胞测序数据质控

使用jupyter平台连接服务器进行R语言和python分析

conda activate jupyter

jupyter notebook --no-browser

2.1 使用python标注双细胞

这里使用Scrublet方法，基于模拟双细胞+k近邻分类的算法，用于检测单细胞 RNA 测序数据中的技术性双细胞。

复制代码

import numpy as np
import pandas as pd
import scanpy as sc

读取原始测序文件

复制代码

ON_brain <- "/storage/public/yongpeng/work/cellranger/ON_cellranger/outs/ON_filtered_feature_bc_matrix.h5"
TB_brain <- "/storage/public/yongpeng/work/cellranger/TB_cellranger/outs/TB_filtered_feature_bc_matrix.h5"

adata_ON = sc.read_10x_mtx(ON_brain,var_names='gene_ids',cache=True)
adata_ON

AnnData object with n_obs × n_vars = 12947 × 32125

var: 'gene_symbols', 'feature_types'

复制代码

adata_TB = sc.read_10x_mtx(TB_brain,var_names='gene_ids',cache=True)

检测并标记出单细胞数据中可能由技术误差产生的双细胞

复制代码

sc.pp.scrublet(adata_ON)
sc.pp.scrublet(adata_TB)

sc.pl.scrublet_score_distribution(adata_ON)
sc.pl.scrublet_score_distribution(adata_TB)

adata_ON 样本的上样浓度较低，预期双细胞率低，因此用更低的阈值（0.16）可能仍能保持较低假阳性；而 adata_TB 样本上样浓度高，双细胞率高，需要更高阈值（0.2）来平衡假阳性和假阴性。

复制代码

sc.pp.scrublet(adata_ON,threshold=0.16)
sc.pp.scrublet(adata_CH,threshold=0.2)

对双细胞进行标记方便后续更具标记删除双细胞

复制代码

adata_TB = adata_TB[~adata_TB.obs['predicted_doublet']]
adata_ON = adata_ON[~adata_ON.obs['predicted_doublet']]

sc.pp.filter_cells(adata_TB, min_genes=100)
sc.pp.filter_genes(adata_TB, min_cells=3)
sc.pp.filter_cells(adata_ON, min_genes=100)
sc.pp.filter_genes(adata_ON, min_cells=3)

讲标记好的文件保存用以后续筛选

复制代码

adata_TB_bc_df = pd.DataFrame(adata_TB.obs.index,columns=['barcode'])
adata_TB_bc_df.to_csv('./adata_TB_barcodes.csv',index=False)

adata_TB_bc_df = pd.DataFrame(adata_TB.obs.index,columns=['barcode'])
adata_TB_bc_df.to_csv('./adata_ON_barcodes.csv',index=False)

得到双细胞标注文件

这里开始转换成R语言

除了去除双细胞外，还要再这一步替换TB中原基因为ON的同源基因以便后续分析

复制代码

library(dplyr)
library(Seurat)
library(patchwork)
library(ggplot2)
library(data.table)
library("Matrix")

Attaching package: 'dplyr'
The following objects are masked from 'package:stats':
filter, lag
The following objects are masked from 'package:base':
intersect, setdiff, setequal, union
Warning message:
"package 'Seurat' was built under R version 4.3.3"
Loading required package: SeuratObject
Warning message:
"package 'SeuratObject' was built under R version 4.3.3"
Loading required package: sp
Attaching package: 'SeuratObject'
The following objects are masked from 'package:base':
intersect, t
Attaching package: 'data.table'
The following objects are masked from 'package:dplyr':
between, first, last

复制代码

ON_brain <- "/storage/public/yongpeng/work/cellranger/ON_cellranger/outs/ON_filtered_feature_bc_matrix.h5"
TB_brain <- "/storage/public/yongpeng/work/cellranger/TB_cellranger/outs/TB_filtered_feature_bc_matrix.h5"

names(ON_dir) <- c("ON_brain")
names(TB_dir) <- c("TB_brain")

ct <- Read10X_h5(ON_dir, use.names = F)
if(inherits(ct, "list")){
    ct <- ct$`Gene Expression`
}
colnames(ct) <- paste(names(ON_dir), colnames(ct), sep = "_")
TB_ct <- dat_combined_Ti <- dsCombineDGE(sceList)

ct <- Read10X_h5(TB_dir, use.names = F)
if(inherits(ct, "list")){
    ct <- ct$`Gene Expression`
}
colnames(ct) <- paste(names(TB_dir), colnames(ct), sep = "_")
TB_ct <- ct

提前准备后续跨物种整合做准备

注释TB基因为ON同源基因

复制代码

oldnames <- data.frame(TB_ID = rownames(TB_ct))
writeLines(oldnames, "/storage/public/yongpeng/work/H5/oldnames.txt")

library(readxl)
library(openxlsx)

TB_to_ON <- read_excel("TB_to_ON.xlsx")

names(TB_to_ON) <- c("ON", "TB")

head(TB_to_ON)

A tibble: 6 × 2

复制代码

newname <- merge(oldnames, TB_to_ON, by = "TB", all.x = TRUE, sort = FALSE)
# 用 match 确保顺序
newname <- newname[match(oldnames$TB, newname$TB), ]

rownames(TB_ct) <- newname

saveRDS(TB_ct,file='/storage/public/yongpeng/work/H5/H5.ON_brain.rds')
saveRDS(TB_ct,file='/storage/public/yongpeng/work/HE/H5.TB_brain.rds')

sc_ON <- CreateSeuratObject(counts = ON_ct,min.cells = 3,min.features = 200)
sc_Tb <- CreateSeuratObject(counts = TB_ct,min.cells = 3,min.features = 200)
#需要根据自己的数据调整参数

saveRDS(sc_ON,file='/storage/public/yongpeng/work/cellranger/ON_brain.rds')
saveRDS(sc_Tb,file='/storage/public/yongpeng/work/cellranger/TB_brain.rds')

sc_ON$orig.ident <- "ON"
sc_Tb$orig.ident <- "TB"

ON_bc <- read.csv('adata_ON_barcodes.csv')
TB_bc <- read.csv('adata_TB_barcodes.csv')

barcodes_to_keep <- ON_bc$barcode
sc_ON$barcode <- colnames(sc_ON)
ON_filtered <- subset(sc_ON,subset = ON_bc$barcode %in% barcodes_to_keep)

barcodes_to_keep <- TB_bc$barcode
sc_Tb$barcode <- colnames(sc_Tb)
TB_filtered <- subset(sc_Tb,subset = sc_Tb$barcode %in% barcodes_to_keep)

saveRDS(TB_filtered,file = "TB.doublefiltered.rds")
saveRDS(ON_filtered,file = "ON.doublefiltered.rds")

至此得到去除双细胞的seurat文件

2.2 QC质控

ON数据质控

复制代码

# 修改active.ident
new_identity <- rep("ON-brain", length(ON_filtered@active.ident))
new_identity <- factor(new_identity)
ON_filtered@active.ident <- new_identity
ON_filtered@active.ident <- factor(ON_filtered@active.ident)
names(ON_filtered@active.ident) <- colnames(ON_filtered)
print(dim(ON_filtered))
mt.genes <- rownames(ON_filtered)[grepl("mt-", rownames(ON_filtered))]
ON_filtered[["percent.mt"]] <- PercentageFeatureSet(ON_filtered, features = intersect(rownames(ON_filtered),mt.genes))

单细胞转录组跨物种比对

一、写在前面

二、实操流程

1.准备工作

1.1 安装cellranger

1.2 准备单细胞测序数据

1.3 下载基因组文件和注释文件

1.4 运行cellranger

1.5 注释同源基因

2.单细胞测序数据质控

2.1 使用python标注 双细胞

2.2 QC质控

2.1 使用python标注双细胞