nmabbasi
ABBASI
Recently Published
Wilcox Analysis RNA Top Markers
NewUmap-Harmony
RNA assay used
NK Proliferating vs Control_filtred_on_mean
UP_genes <- 100
Down_genes <- 100
logFC_up_threshold <- 4
logFC_down_threshold < -4
HSPC vs Control_filtred_on_mean
UP_genes <- 100
Down_genes <- 100
logFC_up_threshold <- 4 # Upregulated logFC threshold
logFC_down_threshold <- -4 # Downregulated logFC threshold
P1_vs_P3_Enrichment
UP → 300 > 1.5
DOWN → 300 < -1.5
P2_vs_P3_Enrichment
UP → 200 > 1.5
DOWN → 150 < -1.5
P1 vs P2 Enrichment Final after Pseudobulk
UP → 200 > 1.5
DOWN → 300 < -1.5
FGSEA- of Malignant CD4Tcells vs Control(Normal CD4 Tcells)
After Pseudobulk approach
Pseudo_bulk_Enrichment_Final
Top200 UP
100DOWN
p_val_adj upregulated gene: 1.193363e-10
downregulated gene: 5.980389e-06
Pseudo_bulk_Enrichment_Final
Top300 UP and DOWN
Number of upregulated genes selected: 300
p_val_adj upregulated gene: 1.751475e-06
downregulated gene: 0.02051237
Pseudo_bulk_Enrichment_Final
Top200 UP and DOWN
upregulated gene: 1.193363e-10
downregulated gene: 0.002239902
PseudoBulk_P2_vs_P3
filtered on mean 1,1,0.05
PseudoBulk_P1_vs_P3
filtered on mean 1,1,0.05
PseudoBulk_P1_vs_P2
filtered on mean
1,1,0.05
PseudoBulk Analysis using Libra_filtred_Mean
Deseq2-LRT_on_list_filtred_on_mean
1.5
-1
0.05
Signature Markers(Malignat_vs_Normal_CD4Tcells) Top30up,down
filtered on MeanExpression
PseudoBulk Analysis using Libra RNA assay-Deseq2-LRT
Audrey Discussion
logFC_up_threshold <- 1
logFC_down_threshold <- -1
pval_threshold <- 1e-4
PseudoBulk Analysis using Libra RNA assay-Deseq2-LRT-Final-P1
logFC_up_threshold <- 1
logFC_down_threshold <- -1
pval_threshold <- 0.05
PseudoBulk Analysis_good_script
Cell_type/CD4T
Wilcox-SCT-Newumap-DE Analysis
min.pct and log2fc=0
after_filtering_on_MeanExpression
Malignant CD4Tcells vs Control(Normal CD4 Tcells)
P1 vs P2
Previous one is with P1 vs P3 and i removed that mistake
adding Mean Expression_P1_vs_P3 - Filtering and Visualization_Patients
P1 vs P3
Added mean expression and filtered on mean expression
Gene Enrichment Analysis (P1_vs_P2)
NewUMAP
MAST_SCT_CLUSTERS used
fgsea-L1 vs Control(Normal CD4 Tcells)
NewUMAP
L1 vs normal
Malignant CD4Tcells vs Control(Normal CD4 Tcells)-Enrichment_final-after_filtering
Its done on all genes with mean expression but we use filters for
avg_log2fc >2 for up <0.05
and
avg_log2fc <-0.5 for down <0.05
Differential Expression Analysis of Malignant CD4Tcells vs Control(Normal CD4 Tcells)-GSEA-Selected_top_genes
Selected_Genes
500 UP
121 Down
Enrichment Analysis
L1 vs Control(Normal CD4 Tcells)-GSEA
NewUMAP
SCT
MAST
Differential Expression Analysis of Malignant CD4Tcells vs Control(Normal CD4 Tcells)-GSEA
GSEA on new UMAP
Different Enrichment Plots
Differential Expression Analysis of Malignant CD4Tcells vs Control(Normal CD4 Tcells)
NewUMAP-fgsea
MAST_SCT
Malignant vs Normal
Filtering and Visualization(cell lines vs Control)
Filtering on newUMAP after running cell lines vs Control
Cell lines vs Normal Control-DE-NewUMAP
L1-L6 vs Normal Control
min.pct=0.01
Filtering and Visualization(MAST, LR and negbinom on SCT)
I used MAST, LR and negbinom
MAST and negbinom gave quite similar significant genes
Filtering and Visualization(MAST on SCT and RNA) on cell_line and default_minpct_logfc
MAST
Default setting
cell_line
Filtering and Visualization(MAST on SCT and RNA)
Filtering of P value <0.05 applied
but we have known genes with p value of 0
MAST on SCT and RNA assay and its comparison to see which is better with default settings
Its using deafult min.pct and logfcT
It give us less genes the using min.pct and lof2fc=0
Filtering and Visualization(MAST on SCT and RNA
I did further filtering for fc T:1
p value:0.05
Filtering and Visualization(MAST on SCT and RNA
Filtered on Mean expression
Alot of genes have p value 1 an important genes
MAST on SCT and RNA assay and its comparison to see which is better
min.pct = 0,
logfc.threshold = 0
Sézary Syndrome Cell Line Analysis
MAST with cell_line ident
L1 to L7 comparison
VolcanoPlot_Malignant_vs_normal
Template for volcanoPlots
Malignat vs Normal
Sézary Syndrome Cell Line derived from each patient DE comparison
Same but with Top10 kegg pathways
Sézary Syndrome Cell Line derived from each patient DE comparison
All Patient based comparison based on clusters
Enrichment analysis included
Sézary Syndrome Cell Line derived from each patient DE comparison
used_patient_origin
NEwumap
Seem like if you dont use harmony clusters batch effect is evident in volcano plot
Sézary Syndrome Cell Line derived from each patient DE comparison
NewUMAP
P1vsP2, P1vsP3, P2vsP3
ClusterBased
Differential Expression Analysis - Filtering and Visualization-NEWUMAP
Malignant_CD4Tcells_vs_Normal_CD4Tcells_RNA_Assay_Wilcox.csv
Malignant_CD4Tcells_vs_Normal_CD4Tcells_SCT_SCTransformed_Wilcox.csv
DE(Malignat_vs_Normal_CD4Tcells) of Harmony Integration
# Define malignant and normal cell lines
malignant_cell_line <- c("L1", "L2", "L3", "L4", "L5", "L6", "L7")
normal_cell_line <- c("PBMC", "PBMC_10x")
DE Analysis Using SCT Assay (SCTransformed Counts)
Next will be: L1 vs healthy
DE Analysis Using RNA Assay (Log-Normalized Counts)
TCR Analysis using harmony integrated NewUMAP with PBMC10x
Hamony integrated UMAP
0.8
with PBMC10x
Different Resolution Tables on harmony integration
Final UMAP
0.5 theta
Different Resolution test on harmony integration
Final UMAP
0.5 theta
Annotated again and removed nonCD4 T cells from Control
Different Resolution Tables on harmony integration
Tables for Resolution
Azimuth annotation done again
Different Resolution test on harmony integration
I annotated the harmony integration again to remove artifacts
Merged All samples with PBMC_10x and removed non CD4 T cells from Control and B cells from L4 regress nCount, percent.mt and rb and apply SCT
didnt remove ILC, NK and CD14 Mono from cell lines and annotated after normalization
use Annotated Robj including PBMC10x to remove NonCD4Tcells from Control and keep just CD4Tcells
28-Jan-
Annotate after SCT
Different Resolution Tables on harmony integration on patient origin and orig.ident-theta-0.5,0.5
patient origin and orig.ident-theta-0.5,0.5
0.4-1.2 tables
Different Resolution test on harmony integration on patient origin and orig.ident-theta-at-0.5,0.5
patient origin and orig.ident-theta-at-0.5,0.5
Different Resolutions: 0.1-1.2
Differential Expression Analysis of Malignant CD4Tcells vs Control(Normal CD4 Tcells)
fgsea_result_kegg <- fgsea(
pathways = kegg_list,
stats = gene_list,
nperm = 1000 # Number of permutations
)
Different evaluation test on harmony integration on patient origin and cell_line-theta-0.5 both
Including clustree
Tables to check distribution
Different evaluation test on harmony integration on patient origin and cell_line-theta-0.5 both
Different Resolution umap
0.1 to 1.2
Harmony integrations of PBMC10x by patient origin and cell_line-theta-0.5 both
Finalized
saved object to:
../0-R_Objects/CD4Tcells_harmony_integrated_0.5_theta_patientorigin_cell_line.Robj
Harmony integrations of PBMC10x by cell_line-theta-0.5
1:15 dim
theta 0.5
cell_line as batch
Merged All samples with PBMC_10x, Removed non CD4 T cells from Control, Apply_SCT_then_Harmony_theta-0.5
1:16
without regressing using SCTransform
use Annotated Robj including PBMC10x to remove NonCD4Tcells from Control and keep just CD4Tcells
I removed B cells from L4
CD14 Mono and ILC and NK just one cell
Differential Expression Analysis - Filtering and Visualization
This analysis includes all genes in SCT
logfc.threshold used = 0
min.pct=0
Then we used added column of mean expression and we used those columns for filtering.
Here you have summary of genes before filtering and after filtering
Differential Expression Analysis - Filtering and Visualization
Its with previous file All genes
min.pct = 0
logfc.threshold = 0
Merged All samples with PBMC_10x, Removed non CD4 T cells from Control, Apply_SCT_then_Harmony_theta-0.5
1:15
Didnt regress anything in SCTransform
Differential Expression Analysis - Filtering and Visualization
Its with previous file
All genes
min.pct = 0
logfc.threshold = 0
Differential Expression Analysis - Filtering and Visualization
Its with previous file
14000 genes
min.pct=default
logfc.threshold=default
Differential Expression Analysis using Harmony Integrated Clusters
heatmap is based on
pvalue and avg_log2FC
Differential Expression Analysis using Harmony Integrated Clusters
celllines vs CD4Tcells normal clusters
min.pct=0
logfc.threshold=0
MAST with batch
Harmony integrations of PBMC10x by cell_line-theta-0.5_removing non CD4Tcells and B cells from L4_also_ILC_NK_CD14_Mono
ILC,NK, CD14 MOno are from L4 where we have B cells so i removed them
Seurat Integration of PBMC10x-HPC-rpca
CD4Tcells
Seurat Integration of PBMC10x-Rserver-rpca-part1
Its done after removal of nonCD4T cells from PBMC
and B cells also from L4
Merged All samples with PBMC_10x and removed non CD4 T cells from Control and B cells from L4 and ILC, NK, CD14 Mono didnt regress nCount and nFeature and apply SCT
I havent regress nCount and nFeature to see UMAP
Merged All samples with PBMC_10x and removed non CD4 T cells from Control and B cells from L4 and ILC, NK, CD14 Mono and regress nCountRNA and nFeatureRNA and apply SCT
1:16
I also removed CD14 Mono
Regress nCount and nFeature
Merged All samples with PBMC_10x and removed non CD4 T cells from Control and B cells from L4 and ILC and NK just one Cell and regress batch and apply SCT
Not a good idea to regress using SCTransform
Harmony integrations of PBMC10x by cell_line-theta-0.5
After removing nonCD4 T cells
use Annotated Robj including PBMC10x to remove ILC and NK-just one Cell
Removed ILC and NK from l2 as it was forming seperate cluster
use Annotated Robj including PBMC10x to remove NonCD4Tcells from Control and B cells from L4
use Annotated Robj including PBMC10x to remove NonCD4Tcells from Control and B cells from L4
Merged All samples with PBMC_10x and removed non CD4 T cells from Control apply SCT
1:22
we got 1:16 by PCA test so I will use that
CD4Tcells in PBMC(Ready to Normalize)
use Annotated Robj including PBMC10x to remove NonCD4Tcells from Control and keep just CD4Tcells
Patients vs PBMC-Tcells
P1,P2,P3
TCR Analysis using harmony integrated UMAP
TCR Analysis using harmony integrated UMAP
Cell Line derived from each patient DE comparison-part2
Top10 categories
Significant genes for strings
0.05
2
-2
Sézary Syndrome Cell Line derived from each patient DE comparison
Cell lines derived from patient
Sézary Syndrome SCpubr Visualization
https://enblacar.github.io/SCpubr-book-v1/04-FeaturePlots.html
Sézary Syndrome Cell Line Top5 gene markers
Top5 and Top10
Differential Expression Analysis of SS vs PBMC10X+PBMC
based on CSV file
Differential Expression Analysis
1vs2
6vs16
InferCNV Analysis
Percentage of cells CNVs
Harmony_Intergrations_and_visualization_of_PBMC-10x.
Its juts to visualize cluster,
clustree and tables
Sézary Syndrome Cell Line Analysis-DE-Integrated
DE integrated
Sézary Syndrome Cell Line Analysis-DE-PBMC10X
DE
Merged with PBMC10x
Differential Expression Analysis of SS vs PBMC10X+PBMC
based it on clusters after harmony integration which I did
Harmony integrations of PBMC10x-part4
final version to Discuss
Multiple Harmony integrations of PBMC10x-part3
on HPC we use sample group, cell line group and cell line
Multiple Harmony integrations of PBMC10x
1:22
Different methods of harmony are tried.
Harmony Integration of PBMC10x-part2
1:22
0.5
Harmony Integration of PBMC10x-Part1
1:22
0.5
RPCA-CCA-Harmony Integration of PBMC10x with NormalizeData-VST on samples part3
Its on Rserver
1:20
1.2
Merged All samples with PBMC_10x and apply SCT on 1:22
Object is saved as All_Samples_Merged_with_10x_Azitmuth_Annotated_SCT_HPC_without_harmony_integration.robj
Harmony Integration of PBMC10x with SCT on samples
old method of normalizing the SCT clusters
Merged All samples with PBMC_10x
1:12 PC
Merged All samples with PBMC_10x
First regressed in SCT for cell_line and then we used harmony
Merged All samples with PBMC_10x
HPC
1:22
SCT normalization of merged samples and Harmony Integration
PBMC10X included
Merged All samples with PBMC_10x and SCT analysis on annotated Object
Did analysis on Rstudio server
Merging all our cell lines and controls(PBMC-PBMC10x) into single seurat object-Robj
Merging all our cell lines and controls(PBMC-PBMC10x) into single seurat object-Robj
PBMC_10x
New Reference
Cytogenetic Analysis
Comparison of inferCNV with Cytogenetics data
TCR Analysis-Part2
New UMAP
WNN analysis of CITE-seq, RNA + ADT_part3
Res=0.9
dims.list = list(1:20, 1:18), modality.weight.name = "RNA.weight"
WNN analysis of CITE-seq, RNA + ADT part2
dims.list = list(1:20, 1:18), modality.weight.name = "RNA.weight"
UMAP of T cells without other PBMC cells using clusters and PC-1:20
UMAP of T cells without other
PBMC cells using clusters and PC-1:20
L1_Merged_first_HPC_PC-1:21-6-old_script
L1_Merged_first_HPC_PC-1:21-6-old_script
L1_Merged_first_HPC_PC-1:50-5
L1_Merged_first_HPC_PC-1:50-5
L1_Merged_first_HPC_PC-1:50-4
L1_Merged_first_HPC_PC-1:50-4
L1_Merged_first_HPC_PC-1:50-3
L1_Merged_first_HPC_PC-1:50-3
UMAP of T cells-PC-1:50-2
Old Script used
Just T cells Analysis_PC-1:50
UMAP of T cells without other PBMC cells using clusters and PC-1:50
L7
Cell Line L7 Analysis
L6
Cell Line L6 Analysis
L5
Cell Line L5 Analysis
L3
Cell Line L3 Analysis
L4
Cell Line L4 Analysis
L4_notebook
Cell Line L4 Analysis
L3_Notebook
Cell Line L3 Analysis
L2_Notebook
Cell Line L2 Analysis
L2
Cell Line L2 Analysis
L1_notebook
Cell Line L1 Analysis
L1
Cell Line L1 Analysis
cell-cell communication using CellChat
Inference and analysis of cell-cell communication using CellChat
Document-Harmony-Integration
1:13
0.1-1.2
Integration by Harmony_on_SCTransform_DATA
Same parameters
Without findNeigbors and FindClusters
1-Harmony Integration_on_SCTransform
Its done on SCTransform data with 1:13 PCA
0.5 Res
Integration by Harmony_by_K_1-50
1:50
0.5
Integration by Harmony_by_K
1;20
0.5
Integration by CCA_by_K_1-12
used K code 1:20 log Norm 1:12 integration
Integration by CCA_by_K_1-20
used K code 1:20 log Norm 1:20 integration
Integration by CCA_by_K
used K code
1:20 log Norm
1:50 integration
CCA_harmony_0.5
dims: 1:15
Resolution Test
0.3-2
Document
Escape Visualization
Document
Analysis of TCR-SS