Put in a gene to see 2D/3D tSNE plots
P6 mouse tibia 10x cells (Figure 1E)
(genotype: C10Cre;Irx3
+/ΔHC
Irx5
+/-
; Rosa26
LSL-tdTomato/+
Het)
Go Up
Go Back
Re-load page
Select a cell set:
chondro-osteogenic cells in Figure 1E
2D or 3D?
2D
3D
with violin plot?
no
linear all
log all
linear pos-only
log pos-only
tSNE or PCA?
tSNE
PCA
UMAP
SCT->CCA
SCT->tSNE
SCT->UMAP
Cell grouping?
By cell pop names (K=8)
By cell gender (K=2)
By CRE_status (K=2)
By CREERT2_status (K=2)
By TDT_status (K=2)
By WPRE_status (K=2)
By TDTWPRE_status (K=2)
TDT_or_Tagln (K=2)
TDT_or_Tagln/Acta2(K=4)
TDT_or_Tagln/Acta2/cell-source (K=16)
Tagln (K=2)
Acta2 (K=2)
T (K=2)
kmeans_1_clusters (K=1)
kmeans_2_clusters (K=2)
kmeans_3_clusters (K=3)
kmeans_4_clusters (K=4)
kmeans_5_clusters (K=5)
kmeans_6_clusters (K=6)
kmeans_7_clusters (K=7)
kmeans_8_clusters (K=8)
kmeans_9_clusters (K=9)
kmeans_10_clusters (K=10)
Lock to color codes
Color codes?
By cell pop names (K=8)
By cell gender (K=2)
By CRE_status (K=2)
By CREERT2_status (K=2)
By TDT_status (K=2)
By WPRE_status (K=2)
By TDTWPRE_status (K=2)
TDT_or_Tagln (K=2)
TDT_or_Tagln/Acta2(K=4)
TDT_or_Tagln/Acta2/cell-source (K=16)
Tagln (K=2)
Acta2 (K=2)
T (K=2)
kmeans_1_clusters (K=1)
kmeans_2_clusters (K=2)
kmeans_3_clusters (K=3)
kmeans_4_clusters (K=4)
kmeans_5_clusters (K=5)
kmeans_6_clusters (K=6)
kmeans_7_clusters (K=7)
kmeans_8_clusters (K=8)
kmeans_9_clusters (K=9)
kmeans_10_clusters (K=10)
Color scales?
Grey -> Red
Grey -> Blues
White -> Greens
White -> Greys
Blue -> Red
Yellow -> Red
Yellow -> Blue
Viridis -> Blue
Red -> Blue
Rainbow
Portland
Picnic
Hot
Electric
Earth
Blackbody
log(expr)
linear
Pos cells on top
manual max
Point size? (
2
)
Point size for zero values? (
3
)
By genes
By markers
By cell properties
e.g. CD24, MGP, GREM1
Add a secondary gene
step1: select a public/in-house dataset
Tiffany's mouse data (n=3+3)
Wilson's ioAF data (n=3)
Ron's mESC data (n=6)
Charles' pulse-chase scRNAseq data (group mean, n=4+1)
Kwok's mouse HCOB scRNAseq data (group mean, n=3)
Peter's mouse ENCC scRNAseq data (group mean, n=6)
Anita Chan's human NP/AF data (n=2+2+2+2)
DS_CDS Illumina Bulk (n=4+5)
Dr QZ Lian's human ESC NP/NCC culture (n=19)
UPenn's mouse data (n=4+4)
Beijing 2013 Chordoma data (n=3+3)
Judith_Hoyland's 7 human embryos (2018, n=5+2)
Bertie and Marioni's mouse E8.25 (pop-mean, n=20)
ENCODE's human total RNAseq data
ENCODE's mouse total RNAseq data (n=220)
single sample
one sample minus another
Max A Posterior (MAP)
step3: select a similarity metric
'
overlap ratio top N genes in each sample (Jaccard index)
overlap ratio top N genes in each sample (Jaccard index), and remove housekeeping genes
Spearman correlation genome-wide
Spearman correlation genome-wide, and remove housekeeping genes
step4: select a geneset
Top 50 genes
Top 100 genes
Top 200 genes
Top 300 genes
Top 400 genes
Top 500 genes
Top 1000 genes
step5: select a normalization of correlations
'
Scaled by sqrt(num of protein coding genes)
Scaled by num of protein coding genes
Scaled by nothing
step6: change colorcodes to pos/neg?
'
No, do not change
Yes, change pos/neg similarity to red/blue
Cutoff:
Only first valid pathway will be shown.
Cutoff:
The all-positive (or your specified) cells will be shown
(To specify a neg marker: add a neg sign before gene.)
--choose a prev saved markerset--
-or-
Also save this set
Cutoff:
choose a dot-shape scheme:
show all-pos (or your specified) cells (as x) only
show all-pos (or your specified) cells as x, others as o; size depends on # pos
show all-pos (or your specified) cells as x, others as o; size depends on level
Scaled or not scaled by num gene expressed per cell?
absolute count (=not scaled)
scaled by gene number per cell
choose a color scheme for all-pos:
keep colors
orange for all-pos, grey otherwise
Multiple pathways are accepted. Union of pathway genes taken.
Over 500 genes could be slow.
Cutoff:
 
How to evaluate pathway activity?
absolute sum of expr levels (=not scaled)
sum of expr levels, scaled by gene # per cell
absolute count of expressed genes (=not scaled)
count of expressed genes, scaled by gene number per cell
 
# of all genes, per cell
# of protein coding genes expressed, per cell
# of processed_pseudogene expressed, per cell
# of lincRNAs genes expressed, per cell
Total # of UMI per cell
Max gene (in UMIs) per cell
Max gene / Total #
TDT levels
WPRE levels
CRE levels
CRE-ERT2 levels
tdTomato levels (Based on num of reads for WPRE4)
Cutoff:
Submit