Gene regulatory network construction
Paired Motif Enrichment
Automatic cell-type annotation
Cell type decomposition of bulk samples
Pseudotime trajectory inference
Explore in Built-in Shiny APP
Gene regulatory networks (GRNs), consisting of the interaction
between transcription factors (TFs) and their target genes, orchestrate
cell-specific gene expression patterns and in turn determine the
function of the cell. To construct cell-type specific GRNs using
single-cell data, we have incorporated a widely used single-cell GRN
analysis tool SCENIC (
Aibar et al.,
2017
) and a plant specific GRN inference tool MINI-EX (
Ferrari et al.,
2022
) into the
scPlant
pipeline.
Construct GRNs using SCENIC
We have incorporated the widely used single-cell GRN analysis tool
SCENIC (
Aibar et
al., 2017
) into the
scPlant
pipeline, and prepared
required cisTarget databases in order to support SCENIC for data
analysis in plants. What’s more, we have provided various visualization
tools to display single-cell GRN results in different ways. Three plant
species (
Arabidopsis thaliana
,
Oryza sativa
,
Zea
mays
) are supported for now, and more plant species will be
supported soon.
Download databases and scripts
The required cisTarget databases and scripts to construct gene
regulatory network can be downloaded
here
.
For other species and genome version
Provided cisTarget databases were built based on the following
species and genome version:
Arabidopsis thaliana:
TAIR10
Oryza sativa:
V7.0
Zea mays:
Zm-B73-REFERENCE-NAM-5.0
For other species and genome version, users can build their own
cisTarget databases by following the protocol
create_cisTarget_databases
.
Create tbl file for other species
For other species, users can create their own motif2TF.tbl file by
creating a data frame with the information of the motifs and their
corresponding TFs, which is the most important information in tbl files
and can be downloaded from some databases such as
JASPAR
and
PlantTFDB
.
Here is the code to create a motif2TF.tbl file:
# Assuming motif2TF is a three-column data frame with the information of motifs and their corresponding TFs like this:
# motif TF source
# MP00120 AT1G01250 PlantTFDB
# MP00100 AT1G01260 PlantTFDB
# Then use the code:
motif2TF
<-
motif2TF
%>%
dplyr
::
transmute
(
`#motif_id`
=
motif
, motif_name
=
motif
, motif_description
=
TF
,
source_name
=
source
, source_version
=
1.1
, gene_name
=
TF
,
motif_similarity_qvalue
=
0.000000
, similar_motif_id
=
"None"
,
similar_motif_description
=
"None"
, orthologous_identity
=
1.000000
,
orthologous_gene_name
=
"None"
, orthologous_species
=
"None"
,
description
=
"gene is directly annotated"
)
write.table
(
motif2TF
, file
=
"motif2TF.tbl"
, sep
=
"\t"
, row.names
=
F
, quote
=
F
)
We also provide a real scRNA-seq data of Arabidopsis thaliana (
Zhang
et al., 2019
) as example data, download
here
.
Installation and pre-process
First, we need to download
MINI-EX
and install the
requirements.
To make MINI-EX work properly,
Nextflow
and
Singularity
need to be installed. MINI-EX offers a Docker container, but if it
doesn’t work, you can create a conda environment and install the
requirements
manually.
Second, we need to produce the input files for MINI-EX.
We provided a script used to produce the input files for MINI-EX,
download
here
.
We provided a real scRNA-seq data of Arabidopsis thaliana (
Zhang
et al., 2019
) as example data, download
here
.
Then, modify the
miniex.config
and change the paths to your
own input files generated by the script above. You can modify the
parameters
according to your particular needs.
Load the Seurat object and specify the path of output files:
SeuratObj
<-
readRDS
(
"SeuratObj.rds"
)
MINIEXouputPath
<-
"/your_path/regulons_output"
Heatmaps showing cluster enrichment and TF expression of each regulon
in each cluster.
enrich_exp_hmp
(
SeuratObj
,
MINIEXouputPath
)
Dimension reduction plot showing cluster enrichment and TF
expression.
enrich_exp_scatter
(
SeuratObj
,
MINIEXouputPath
, gene
=
'AT1G75390'
)
Network diagram showing top regulons of each cluster according to
Borda ranking.
topRegulons_MINIEX
(
MINIEXouputPath
, topn
=
5
)
Network diagram showing targets of each regulon.
targets_MINIEX
(
MINIEXouputPath
, cluster
=
23
, regulons
=
"AT3G16770"
)
Dot plot showing top regulons of each cluster according to Borda
ranking.
BordaRank_MINIEX
(
MINIEXouputPath
, cluster
=
23
, topn
=
10
)
