Gene expression values for individual genes for each cell line were tested for correlation with cell-line rbs9.6, an indicated of R-loop levels.
Many datasets were not normally distributed and therefore did not meet the requirements of Pearson correlation. Correlations were instead all performed using Spearman so that the results could be more directly compared between genes. The genes with a significant gene expression correlation (p < 0.05) to R-loop values were considered for further analysis.
The file containing these genes is loaded:
#Load gene expression file from gene expression output
pos_gene_express <- read.csv("Output/Outputdata/geneexpressionRlooppos.csv")
neg_gene_express <- read.csv("Output/Outputdata/geneexpressionRloopneg.csv", header = TRUE)These are the top negatively correlated genes for gene expression, ordered by correlation estimate.
These are the top positively correlated genes for gene expression, ordered by correlation estimate.
Enrichr is employed for genes with negatively correlated expression and positively correlated expression to rbS9.6, respectively.
There are about 19,117 genes that obtained a p-value (not NA) for the correlation, and 3,263 of these obtained a p-value below 0.05 (not adjusted). Enrichr was used to determine which databases would be of most interest, mainly from their adjusted p-values and combined score, then the enrichr R package was used to load these results (1).
Of the 3263 genes with p < 0.05, 1,946 of these were negatively correlated with R-loop value, suggesting that these are expressed less with increasing R-loop levels. The figure for negatively correlated gene expression can be found below.
A link to the enrichr site for these genes can be found here. https://maayanlab.cloud/Enrichr/enrich?dataset=aecb2bc6c561a3e6ac08467f64f84187
#use enrichr for negatively correlated genes
#select databases of interest based on the enrichr website input
setEnrichrSite("Enrichr") # for human genes ## Connection changed to https://maayanlab.cloud/Enrichr/
## Connection is Live!
websiteLive <- TRUE
dbs <- c("ENCODE_and_ChEA_Consensus_TFs_from_ChIP-X", "ARCHS4_TFs_Coexp", "TF_Perturbations_Followed_by_Expression", "TRRUST_Transcription_Factors_2019", "lncHUB_lncRNA_Co-Expression", "ChEA_2016", "BioPlanet_2019", "WikiPathways_2019_Human", "KEGG_2019_Human", "ARCHS4_Kinases_Coexp", "MSigDB_Hallmark_2020", "Elsevier_Pathway_Collection", "GO_Biological_Process_2018", "GO_Molecular_Function_2018", "GO_Cellular_Component_2018", "Human_Phenotype_Ontology", "COVID-19_Related_Gene_Sets", "DisGeNET", "DSigDB", "ARCHS4_IDG_Coexp", "Rare_Diseases_GeneRIF_ARCHS4_Predictions", "Rare_Diseases_AutoRIF_ARCHS4_Predictions", "CCLE_Proteomics_2020", "Human_Gene_Atlas", "ProteomicsDB_2020", "ARCHS4_Tissues", "ARCHS4_Cell-lines", "ESCAPE", "NCI-60_Cancer_Cell_Lines", "Enrichr_Submissions_TF-Gene_Coocurrence", "Reactome_2016", "BioPlex_2017", "Jensen_DISEASES", "Jensen_COMPARTMENTS", "ClinVar_2019", "DepMap_WG_CRISPR_Screens_Sanger_CellLines_2019", "Rare_Diseases_AutoRIF_Gene_Lists", "Virus_Perturbations_from_GEO_down", "Virus_Perturbations_from_GEO_up", "Allen_Brain_Atlas_10x_scRNA_2021", "NCI-60_Cancer_Cell_Lines", "Disease_Perturbations_from_GEO_up")
#get list of genes from table
neg_genes <- neg_gene_express$unique_genes
neg_genes[1]
if(websiteLive) {
enriched_neg <- enrichr(neg_genes, dbs)
}The top 8 results from each of the databases of interest relating to transcription for the negatively correlated gene set are displayed below.
The top negatively expressed transcrpition genes from ChEA_2016 with highly significant adjust p-values are RELA, CJUN, ATF3, and BRD4, within specific cell lines.
RELA codes for part of the most common NK-KB transcription factor, which activates transcription and is involved in many processes and cell types (1). CJUN is a TF, and can be involved in cAMP signaling pathway stimulation and colorectal cancer (2). ATF3 is also involved in cAMP in the response element (3). BRD4 regulates transcription and is involved in epigenetic memory, but is also binds with RELA for NF-KB activation (4). It has been considered as a target for inhibition in leukemia treatment (4).
1 https://www.uniprot.org/uniprot/Q04206#function
2 https://www.uniprot.org/uniprot/P05412#function
3 https://www.uniprot.org/uniprot/P18847#function
4 https://www.uniprot.org/uniprot/O60885#function
The top negatively expressed transcription genes from ChEA_2016 and ENCODE consensus with significant p-values are NFE2L2, SUZ12, and IRF8. NFE2L2 is involved in autophagy and oxidative stress, and is allowed to accumulate during oxidative stress instead of being broken down by the BCR(KEAP1) complex (1).
SUZ12 is part of the PRC2 complex which is involved in repressing transcription, and one of the genes it represses is MYT1 (2). IRF8 can activate or repress transcription, and is part of cell differentiation in CD8+ dendritic cells (WBC) (3).
1 https://www.uniprot.org/uniprot/Q16236#function
2 https://www.uniprot.org/uniprot/Q15022#function
3 https://www.uniprot.org/uniprot/Q02556#function
The top negatively expressed transcrpition genes from ARCHS4 transcription factors consensus with highly significant p-values are FOXC2 (as well as FOXL1 and FOXF1), GLIS1, FBN1, HIF1A, SNAI2, CREB3L1, and PRRX2.
FOX Genes:
FOXC2 activates transcription and it could be part of the formation of special mesenchymal tissues (1). FOXL1 is involved in the GI epithelia profileration and differentiation (2). FOXF1 is likely to activate transcription for lung-related genes (3).
GLIS1 affects transcription of genes related to cell differentiation in that it can stop differentiation of multipotent cells including mesenchymal cells (4).
FBN1 is involved in ECM and bone growth (osteoblasts maturation and osteoclast formation) (5).
HIF1A is a TF particularly involved in the response to hypoxia, and in monocytes in a SARS-COV-2 infection it promotes an inflammatory response (6).
SNAI2 is involved in many processes, including repressing the expression of BRCA2 in breast cells, ITGA3 in epidermal keratinocytes, and CDH1, as well as roles in osteoblasts and mesenchymal stem cells (7).
CREB3L1 is a TF that is involved in bone matrix secretion and formation in osteoblasts, and is also involved in ER stress and the cAMP response element in astrocytes (8).
PRRZ2 could be involved in wound healing during fetal development (9)
1 https://www.uniprot.org/uniprot/Q99958#function
2 https://www.uniprot.org/uniprot/Q12952#function
3 https://www.uniprot.org/uniprot/Q12946#function
4 https://www.uniprot.org/uniprot/Q8NBF1#function
5 https://www.uniprot.org/uniprot/P35555#function
6 https://www.uniprot.org/uniprot/Q16665#function
7 https://www.uniprot.org/uniprot/O43623#function
8 https://www.uniprot.org/uniprot/Q96BA8#function
9 https://www.uniprot.org/uniprot/Q99811#function
The top negatively expressed transcription genes from TF perturbations consensus with highly significant p-values are MYC (UP), EPAS1 (UP), FOXQ1 (UP), ERG (DOWN), ESR1 (UP), FLI1 (UP), ATF3 (DOWN), ATF6 (UP).
The MYC gene codes for a TF and activates genes related to growth and is involved in angiogenesis via promotion of VEGFA (1). It is related to Burkitt Lymphoma and multiple myeloma (2).
EPAS1 also is involved in VEGF expression, and therefore is also involved in angiogenesis, and also is involved in hypoxia response (3).
FOXQ1 is involved in differention of hair follicles (4).
ERG is a TF whose perturbation is ‘down’ in this case. It is involved in cell development such as hematopoesis, and it is associated with Ewing sarcoma and acute myeloid leukemia (5).
ERS1 codes for a hormone receptor with many functions, some of which are related to NF-KB, CCL2 and IL8 (6).
FLI1 is a TF that is involved in hematopoietic stem cell differentiation, and is related to Ewing sarcoma, acute lymphoblastic leukemia, and other blood disorders (7).
ATF3 is ‘down’ in these results, has already been discussed as being involved in the cAMP reponse element (8).
ATF6 is dependent on cAMP, and is removed from the ER in ER stress (9).
1 https://www.uniprot.org/uniprot/P01106#function
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=MYC
3 https://www.uniprot.org/uniprot/Q99814#function
4 https://www.uniprot.org/uniprot/Q9C009#function
5 https://www.genecards.org/cgi-bin/carddisp.pl?gene=ERG
6 https://www.uniprot.org/uniprot/P03372#function
7 https://www.genecards.org/cgi-bin/carddisp.pl?gene=FLI1
8 https://www.uniprot.org/uniprot/P18847#function
The top negatively expressed long non-coding RNA genes from lncHUB lncRNA coexpressed with highly significant p-values are LINC02376, MSC-AS1, LINC01204, TENM3-AS1, GORAB-AS1, LINC01615, LINC01583, P4HA2-AS1, and LINC00601.
MSC-AS1 is associated with Familial Episodic Pain Syndrome (1), LINC01204 is associated with hereditary hemorrhagic telangiectasia (2), which involves blood vessel abnormalities (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=MSC-AS1
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=LINC01204
3 https://www.malacards.org/card/hereditary_hemorrhagic_telangiectasia
The top negatively expressed gene pathways from BioPlanet 2019 coexpressed with significant p-values are mostly involved in ECM functions. Perhaps of specific interest:
TWEAK, which is involved in apoptosis (1).
Platelet amyloid precursor protein pathways, which is for a protein found in platelets and associated with Alzheimer disease (2).
Integrins in angiogenesis, because several other links to angiogenesis have been found earlier in this report.
1 https://pubmed.ncbi.nlm.nih.gov/9405449/
2 https://pubmed.ncbi.nlm.nih.gov/28611024/
The top negatively expressed gene pathways from Wikipathways 2019 with significant p-values are miRNA targets in ECM and membrane receptors, miR-509-3p alteration of YAP1/ECM (Extracellular matrix related), and Senesence and Autophagy in cancer, and platelet-mediated interactions with vascular and circulating cells.
Again, this brings up more ECM/membrane receptor and platelet relationships with the negatively correlated gene set.
The top negatively expressed gene pathways from KEGG 2019 Human with significant p-values are focal adhesion, phagosome, and ECM-receptor interaction. This again suggests more ECM relationships (ex. focal adhesion, ECM-receptor interaction), but introduces specifically phagosomes and lysosomes. There is also more discussion of the ER protein processing, which can be involved in ER stress (1).
1 https://www.genome.jp/kegg-bin/show_pathway?ko04141#:~:text=KEGG%20PATHWAY%3A%20Protein%20processing%20in%20endoplasmic%20reticulum&text=The%20endoplasmic%20reticulum%20(ER)%20is,the%20help%20of%20lumenal%20chaperones.&text=Accumulation%20of%20misfolded%20proteins%20in,unfolded%20protein%20response%20(UPR).
The top negatively expressed gene pathways from ARCHS4 kinases with highly significant p-values are NEK7, MYLK, DDR2, PDGFRB (and PDGFRA), and ABL2. NEK7 kinase is important in mitosis (1).
MYLK is involved in smooth muscle contraction, but has myriad roles in wound healing, breast cancer, cervical cancer, asthma and more (2).
DDR2 is involved in ECM remodelling, cutaneous wound healing, and bone development (3).
PDGFRB along with PDGFRA creates a tyrosine-protein kinase receptor involved in skeletal and blood vessel development, wound healing, and bone marrow mesenchymal stem cell differentiation (4,5).
ABL2 is involved in a number of cell adhesion, motility, and actin cytoskeleton (6). It is also associated with leukemia, including chronic myeloid leukemia (7).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=NEK7
2 https://www.uniprot.org/uniprot/Q15746#function
3 https://www.uniprot.org/uniprot/Q16832#function
4 https://www.uniprot.org/uniprot/P09619#function
5 https://www.genecards.org/cgi-bin/carddisp.pl?gene=PDGFRA
6 https://www.uniprot.org/uniprot/P42684#function
7 https://www.genecards.org/cgi-bin/carddisp.pl?gene=ABL2
The top negatively expressed gene pathway from MSigDB Hallmark 2020 with highly significant p-values is by far the epithelial mesenchymal transition (1). This pathway is involved in wound healing (similar to previously seen). However, NF-KB and hypoxia have also been discussed previously.
1 https://www.gsea-msigdb.org/gsea/msigdb/cards/HALLMARK_EPITHELIAL_MESENCHYMAL_TRANSITION.html
The top negatively expressed gene pathway from Elsevier pathway collection with significant p-values is by far genes with mutations in cancer immune escape. Immune escape functions of cancer are when the cancer modulates the immune response in some way like gene mutations or epigenetic changes (1).Two of the other pathways in this list also involve cancer immune escape.
## [1] 4
Again, more ECM relationships are described in the ontologies referring to antigen presentation and ECM organisation.
A new topic of an opic-related ontology can be found in the retinal ganglion cell axon guidance.
The top negatively expressed gene ontologies from GO molecular function 2018 with significant p-values are vinculin binding and platelet-derived growth factor binding.
Vinculin is involved in embryonic development and cell adhesion (1).
Platelet-derived growth factor (PDGF) has already been discussed in the context of the kinase receptors for it, and it is involved in development of various tissues and wound healing.
Collagen-binding and integrin binding are again probably involved in the ECM.
1 Peng et al.(2011) New insights into Vinculin function and regulation. International review of cell and molecular biology. https://doi.org/10.1016/B978-0-12-386043-9.00005-0
The top negatively expressed gene ontologies from GO cellular component 2018 with significant p-values are endoplasmic reticulum lumen, focal adhesion, and phagocytic vesicle membrane.
The ER lumen ontology may be related to the ER stress mechanisms already discuessed. Focal adhesion and phagocytic vesicle membrane again related to the ECM and phagosomes.
The top negatively expressed genes relation to diseases/drugs from COVID-10 related gene sets with highly significant p-values are related to Calu-3, and SARS-CoV perturbation Up Genes bronchial epithelial 2B4. Calu-3 and A549 are both lung adenocarcinomas (1,2).
1 https://web.expasy.org/cellosaurus/CVCL_0609
2 https://web.expasy.org/cellosaurus/CVCL_0023
The top negatively expressed genes relation to diseases/drugs from DSigDB were mostly histamine receptor antagonists.
Suloctidil was a calcium agonist which altered blood platelets, but causes hepatotoxicity (1). Pizotifen is a diarylmethane used in migrane treatment by acting as a serotonin agonist and is a histamine antagonist (2). Terfenadine is a diarylmethane prodrug of a histamine H1-receptor antagonist (3). Prenylamine is also diarylmethane formerly used for angina pectoris (4). Chloropyramine is also an antihistamine (5). Triprolidine and promethazine are also an antihistamines (6,7).
1 https://drugs.ncats.io/substance/XV1N1XY17K
2 https://pubchem.ncbi.nlm.nih.gov/compound/Pizotifen
3 https://pubchem.ncbi.nlm.nih.gov/compound/5405
4 https://pubchem.ncbi.nlm.nih.gov/compound/9801
5 https://pubchem.ncbi.nlm.nih.gov/compound/25295
6 https://pubchem.ncbi.nlm.nih.gov/compound/5282443
7 https://pubchem.ncbi.nlm.nih.gov/compound/4927
The top negatively expressed genes relation to diseases/drugs from ARCHS4 IDG Coexpression with significant p-values are mainly related to NEK7 IDG kinase. NEK7 has already been discussed as being an important gene for mitosis.
The top negatively expressed genes relation to diseases/drugs from Rare Disease GeneRIF ARCHS4 predications are nephrogenic systemic fibrosis, and more with highly significant adjusted p-values and high combined scores.
Osteolysis syndrome is a very rare disease which includes resorption of the distal limb and lack of certain phalanges (1).
POEMS syndrome is a rare disorder which involves Polyneuropathy, Organomegaly, Endicrinopathy, Monoclonal gammopathy and Skin changes, but notably VEGF may be involved in it (2).
Congenital chylothorax is a rare condition in neonates in which chyle builds up in the pleural space, and can be lymphatic sytem related (3).
Familial platelet disorder with associated myeloid malignancy is a rare disorder linked to RUNX1/AML1/CBFA2 mutations, and is linked to increased risk of acute myeloid leukemia (4).
Nephrogenic systemic fibrosis is again a rare disease in which there is fibrosis of the skin and subcutaneous tissue after exposure to IV contrast containing gadolinium in someone with reduced kindey function (5).
1 https://www.orpha.net/consor/cgi-bin/OC_Exp.php?lng=EN&Expert=2776
2 https://rarediseases.info.nih.gov/diseases/7411/poems-syndrome
3 https://www.orpha.net/consor/cgi-bin/OC_Exp.php?lng=en&Expert=264688
4 https://rarediseases.org/rare-diseases/familial-platelet-disorder-with-associated-myeloid-malignancy/#:~:text=Familial%20platelet%20disorder%20with%20associated%20myeloid%20malignancy%20(FPD%2FAML),known%20as%20AML1%20or%20CBFA2.
5 https://rarediseases.org/rare-diseases/nephrogenic-systemic-fibrosis/
The top negatively expressed genes relation to diseases/drugs from Rare Diseases AUTORIF ARCHS4 predications with highly significant p-values are Ptosis coloba mental retardation, Quebec platelet disorder, lathyrism, malignant teratocarcinosarcoma, Bruck syndrome 1 and more. Quebec platelet disorder causes increased bleeding and is a platelet granule disorder (1).
Lathyrism is a rare disease which causes paralysis after eating too mean types of a type of pea (2).
Bruck syndrome 1 is a rare syndrome which include poor bone development, and was mapped to ch17p12 (3).
1 https://www.orpha.net/consor/cgi-bin/OC_Exp.php?lng=EN&Expert=220436
2 Spencer PS, Schaumburg HH. Lathyrism: a neurotoxic disease. Neurobehav Toxicol Teratol. 1983 Nov-Dec;5(6):625-9. PMID: 6422318.
3 https://rarediseases.info.nih.gov/diseases/1029/bruck-syndrome-1
## [1] 7
The top negatively expressed genes related to cell lines from Top CCLE Proteomics 2020 with highly significant p-values are mostly CNS (four glioblastoma) and some Kidney and lung. U118MG is an astrocytoma cell line (1).
OSCR2 is a renal cell carcinoma line (2).
U87MG is probably a glioblastoma cell line (although there are issues with its identification) (3). KNS81 is also a glioblastoma (4).
NCIH2052 is a pleural mesothelioma cell line (5).
A172 and SNU1105 are also a glioblastoma lines (6,7).
1 https://web.expasy.org/cellosaurus/CVCL_0633
2 https://web.expasy.org/cellosaurus/CVCL_1626
3 https://web.expasy.org/cellosaurus/CVCL_0022
4 https://web.expasy.org/cellosaurus/CVCL_2799
5 https://web.expasy.org/cellosaurus/CVCL_1518
6 https://web.expasy.org/cellosaurus/CVCL_0131
Using the top negatively expressed gene the top cell line from the Human Gene Atlasy with highly significant p-values is overwhelmingly smooth muscle.
Using the top negatively expressed gene the top cell lines from the ProteomicsDB 2020 with highly significant p-values is breast, brain, and kidney. The specific cell lines are invasive ductal carcinoma Hs-578T twice (1), glioblastoma SNB-75 twice (2), gliosarcoma SF-539 (3) twice, renal cell carcinoma RXF-393 (4), and another glioblastoma GaMG (5).
1 https://web.expasy.org/cellosaurus/CVCL_0332
2 https://web.expasy.org/cellosaurus/CVCL_1706
3 https://web.expasy.org/cellosaurus/CVCL_1691
4 https://web.expasy.org/cellosaurus/CVCL_1673
5 https://web.expasy.org/cellosaurus/CVCL_1226
Using the top negatively expressed gene the top cell types from ARCHS4 tissues with highly significant p-values are fibroblast, osteoblast, Kupffer cells, vascular smooth muscle, and astrocytes. This makes sense for the the fibroblast, osteoblast, vascular smooth muscle, and astrocytes given the previous results, but there has not been as much indication of Kupffer cells although they are immune cells.
Using the top negatively expressed gene the top cell lines from ARCHS4 cell-lines with highly significant p-values are IMR90, NHDF, BJ CELL, A172, CAKI1, and HUVEC.
IMR90 is a normal fetal lung fibroblast tissue (1). NHDF are normal human dermal fibroblasts from juvenile foreskin or adult skin (2). BJ cell is also a normal human fibroblast line from foreskin (3). A172 is again, a glioblastoma as described earlier.
CAKI1 is a clear cell renal carcinoma (4).
T98G is another glioblastoma (5).
1 https://www.lgcstandards-atcc.org/products/all/CCL-186.aspx?geo_country=se
2 https://www.promocell.com/product/normal-human-dermal-fibroblasts-nhdf/
3 https://www.lgcstandards-atcc.org/products/all/CRL-2522.aspx?geo_country=se
4 https://web.expasy.org/cellosaurus/CVCL_0234
5 https://web.expasy.org/cellosaurus/CVCL_0556
Using the top negatively expressed gene from ESCAPE with highly significant p-values are POU5F1 DOWN and SUZ12 UP. SUZ12 has already been discussed as being part of the PRC2 complex. POU5F1 codes for a neural transcription factor and has been associated with adult acute myeloid leukemias and affects ESR1 TF activity (1).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=POU4F1
Using the top negatively expressed gene the top cell lines from NCI-60 Cancer cell lines with highly significant p-values is by far SNB75, which has been discussed before and is a glioblastoma.
Some specific genes/proteins appear important:
NEK7, NFE2L2, SUZ12, PDGF, and MYT1
Drugs that target histamine H1 receptors may also have some relationship with fewer Rloops.
Specific functions/cell types that appear related have been grouped, although they are also interrelated:
Mitosis
ECM functions
ECM + adhesion + phagosomes + lysosomes + antigen presentation + integrin binding
Apoptosis
Bone and blood
Bone formation/development + bone marrow + hematopoeisis + WBC + platelets
Mesenchyme and differentiation
Mesenchymal tissue + epithelial mesenchymal transition + differentiation
ER stress
ER stress + cAMP response element + hypoxia + ER protein processing
Fibroblast and skin
Wound healing + skin + subcutaneous tissue + fibroblast cell lines + fibrosis + hair follicles
Carcinomas breast + renal + lung
Smooth muscle
Angiogenesis
MYC + VEGFA + angiogenesis
Blood disorders
Blood Ewing sarcoma + leukemia (esp chronic and acute myeloid leukemia) + Burkitt lymphoma
CNS cancers
Glioblastoma cell lines + Astrocytoma cell lines
Lung cancer
Lung adenocarcinoma + pleura issues
Cancer immune escape
Retinal tissue
The figure for positively correlated gene expression can be found below. A link to the enrichr site for these genes can be found here.
https://maayanlab.cloud/Enrichr/enrich?dataset=29b2554bfc9cb523e52737779ba120c4
There was a total of 1,290 positively correlated expression genes (p < 0.05).
#Load gene expression file from gene expression output
length(fewer_pos_gene_express$Gene)
#get list of genes from table
pos_genes <- pos_gene_express$unique_genes
pos_genes[1]
if(websiteLive) {
enriched_pos <- enrichr(pos_genes, dbs) #need to get a list of genes
}
plotdbspos <- function(x) {
enriched_pos[[x]] %>%
top_n(8, Combined.Score) %>%
arrange(desc(Combined.Score)) %>%
mutate(Term = factor(Term, levels = rev(Term))) %>%
ggplot(aes(x = Term, y = Combined.Score, fill = Adjusted.P.value)) +
geom_bar(stat = "identity") +
theme_bw(base_size = 12) +
scale_x_discrete(labels = function(x) stringr::str_wrap(x, width = 20)) +
xlab(NULL) +
labs(title = stringr::str_wrap(paste("Top", x, "pos. express")), #FIX !
fill = "Adjusted P Value") +
ylab("Combined Score") +
ggpubr::rotate()
}The transcription factors from ChEA 2016 related to the positively correlated gene expression set with significant adjusted p-value and highest combined score are MYC and EKLF in mice.
The MYC TF has already been mentioned as being involved growth and angiogenesis via promotion of VEGFA.
EKLF is a TF involved in erythropoiesis, and is linked to anemia (1).
E2F7 is a TF that has many functions in angiogenesis (maybe via VEGFA binding), placental development, and DNA damage response (2). It is linked to retinal cancer (2).
XRN2 is involved in R-loop formation via promotion of RNA polymerase II transcription termination? (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=KLF1
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=E2F7
3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=XRN2
MYC again appears in this database for both ENCODE and ChEA.
KAT2A is a TF involed in numberous functions, including growth and differentiation of various tissues and systems including the heart, limb, craniofacial cartilage and bone, etc. (1) The E2Fs are involved in regulating the cell type and tumor suppressor proteins (2). E2F1 is linked to Retinoblastoma and glioblastoma multiforme (2). E2F6 is linked to retinal cancer and familial retinoblastoma (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=KAT2A
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=E2F1
3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=E2F6
The transcription factors from ARCHS4 TFs coexpression related to the positively correlated gene expression set with significant adjusted p-value and highest combined score are TP53, E2F3, CBX2, MYBL2, TCF3, and PHB2.
TP53 is a tumor suppressor (stops growth or causes apoptosis), and mutations in it are common in cancer (1).
E2F3 is again in the E2F family, and changes in its copy number is common in cancer (2).
CBX2 forms part of a PcG complex, and is involved in differentiation in the embryo and sexual development (3).
MYBL2 is involved in differentiation, proliferation and survival (4).
TCF3 is involved in various tissue differentiation in the embryo, and lymphopoeisis (5).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=TP53
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=E2F3
3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=CBX2
4 https://www.genecards.org/cgi-bin/carddisp.pl?gene=MYBL2
5 https://www.genecards.org/cgi-bin/carddisp.pl?gene=TCF3
The transcription factors from TF perturbations followed by expression related to the positively correlated gene expression set with significant adjusted p-value and highest combined score are GATA6 DOWN, CTCF SINRNA HEK293 UP, EZH" SHRNA PROE DOWN, and YY1 KD UP.
GATA6 is involved in formation of the GI tract, lung and heart and as such mutations are linked to congenital defects (1).
CTCF has several functions, including repressing MYC, but generally it can also act as a activator (2).
EZH2 codes for a PcG protein and is important in embryonic stem cells and possibly the CNS and hematopoiesis (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=GATA6
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=CTCF
3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=EZH2
The transcription factors from lncHUB lncRNA co-expression related to the positively correlated gene expression set with significant adjusted p-value and highest combined score are LINC02067, FOXP4-AS1, LINC00858 and DLGAP1-AS2. FOXP4-AS1 is linked to colorectal cancer (1), while LINC00858 is linked to colorectal cancer and lung cancer (2). DLGAP1-AS2 is linked to astrocytoma and chromosome 18P deletion syndrome (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=FOXP4-AS1 2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=LINC00858 3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=DLGAP1-AS2
The transcription factors from Enrichr submissions TF-gene cooccurrence related to the positively correlated gene expression set with significant adjusted p-value and highest combined score are U2AF1, HMGB1, GTF3A, and SSB.
U2AF1 is involved in splicing (the small subunit) and is associated with a variety of cancers including lung, pancreatic, and acute myeloid leukemia (1).
HMGB1 has many functions, including transcription, DNA organization, differentiation, and more (2).
GTF3A acts in the formation of the ribosomal large subunit (3).
SSB is involved in RNA metabolism (4).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=U2AF1
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=HMGB1
3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=GTF3A
4 https://www.genecards.org/cgi-bin/carddisp.pl?gene=SSB
## [1] 7
Most of these are related to ribosomes and translation. The top pathways that for the positively correlated gene set from BioPlanet 2019 are Cytoplasmic ribsomal proteins which are involved in translation (1), virus related (viral RNA transcription and replication, infection), translation, and Cap-dependent translation initiation.
1 https://www.wikipathways.org/index.php/Pathway:WP477
The top pathway for the positively correlated gene set from WikiPathways 2019 is overwhelmingly cytoplasmic ribosomal proteins.
The top pathway for the positively correlated gene set from KEGG 2019 is overwhelmingly ribosome.
The top human kinases for the positively correlated gene set from ARCHS4 Kinases Coexpression are STK26, CHEK1, and CSNK2A1.
STK26 is involved in growth and apoptosis (1).
CHEK1 is involved in DNA repair and cell-cycle regulation (2).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=STK26
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=CHEK1
The top pathways for the positively correlated gene set from MSigDB Hallmark 2020 are Myc targets V2 (and V1), and E2F targets, which we have already see a lot of in the previous section.
MYC targets V1 and V2 are just two different version of a set of genes regulated by the MYC TF (1).
E2F are targets of E2F TFs.
G2-M checkpoints are genes involved in going through cell division (2).
1 https://www.gsea-msigdb.org/gsea/msigdb/cards/HALLMARK_MYC_TARGETS_V1
2 https://www.gsea-msigdb.org/gsea/msigdb/cards/HALLMARK_G2M_CHECKPOINT
The top pathways for the positively correlated gene set from Reactome 2016 are peptide chain elongation, translation (eukaryotic translation elongation, viral mRNA translation, and eurkaryotic translation termination), selenocysteine synthesis, and Nonsense-mediated decay.
The top pathways for the positively correlated gene set from BioPlex 2017 are RRS1, MRPL24 (and several other MRPLs).
RRS1 is involved in ribosome biogenesis (1).
MRPs are mitochondrial ribosomal proteins (2).
CNBP binds single stranded DNA and RNA (3).
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=RRS1 2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=MRPL24 3 https://www.genecards.org/cgi-bin/carddisp.pl?gene=CNBP
## [1] 6
The ontologies related to the positively correlated gene set in GO Biological Process 2018 were viral related (viral gene expression, viral transcription cotranslational protein targeting to membrane), SRP-dependent cotranslational protein targeting to membrane, and protein targeting to ER.
The ontology related to the positively correlated gene set in GO Molecular function 2018 was predominantly ubiquitin-protein transferase inhibitor activity.
The ontologies related to the positively correlated gene set in GO cellular component 2018 were ribosome related.
The ontologies related to the positively correlated gene set in Human Phenotype Ontology were mostly related to limb hypertonia, but also somewhat related to abnormalities in red blood cells (erythroid lineage cells and macrocytic anemia). Limb hypertonia is when muscle tone is kept stiff due to damage to the nervous system (1).
1 https://www.ninds.nih.gov/Disorders/All-Disorders/Hypertonia-Information-Page#:~:text=Hypertonia%20is%20a%20condition%20in,tell%20the%20muscle%20to%20contract.
The ontologies related to the positively correlated gene set in Jensen DISEASES were mostly related to Diamond-Blackfan anemia.
Diamond-Blackfan anemia is a disorder in which there are insufficent RBCs as well as anomalies in the form of several parts of the body (1).
Meningoencephalitis is an infection of the meninges and brain (2).
1 https://rarediseases.info.nih.gov/diseases/6274/diamond-blackfan-anemia
2 https://www.hopkinsmedicine.org/health/conditions-and-diseases/herpes-hsv1-and-hsv2/herpes-meningoencephalitis
## [1] 7
Again, the only highly related disease to the positively correlated expression genes in ClinVar 2019 is Diamond-Blackfan anemia.
The top highly related disease cell lines to the positively expressed genes in DepMap WG CRISPR Screens Sanger Cell lines 2019 include OCI-AML(2 and 3), SW1573, MHH-ES-1, and U-251MG.
OCI-AML2 and 3 are acute myeloid leukemia, which has come up numerous times in this analysis (1,2).
SW1573 is a lung adenocarcinoma (3).
MHH-ES-1 is Ewing sarcoma (4).
U-251MG is astrocytoma (5).
1 https://web.expasy.org/cellosaurus/CVCL_1844
2 https://web.expasy.org/cellosaurus/CVCL_1619
3 https://web.expasy.org/cellosaurus/CVCL_1720
4 https://web.expasy.org/cellosaurus/CVCL_1411
5 https://web.expasy.org/cellosaurus/CVCL_0021
The top highly related diseases to the positively expressed genes in Rare Diseases AutoRIF Gene Lists are, again, Diamond-Blackfan anemia but also pure red cell aplasia.
Pure red cell aplasia occurs when the bone marrow produces fewer and fewer RBCs (1).
1 https://rarediseases.org/rare-diseases/pure-red-cell-aplasia-acquired/
The top highly related kinases to the positively expressed genes in ARCHS4 IDG coexpression are UCK2 and PKMYT1.
UCK2 codes for a kinase that related to testicular germ cell tumor and distal musclar dystrophy with anterior tibial onset (1).
PKMTY1 (aka MYT1) codes for a kinase involved in cell cycle regulations by inactivating a kinase in the G2/M transition (2). As noted earlier, it is repressed by SUZ12.
1 https://www.genecards.org/cgi-bin/carddisp.pl?gene=UCK2
2 https://www.genecards.org/cgi-bin/carddisp.pl?gene=PKMYT1
The top highly related diseases to the positively expressed genes in DisGeNet (gene disease associations) are Aase Smith syndrome 2, and abnormalities of various sytems.
Aase Smith syndrome 2 is a synonym for Diamond-Blackfan anemia (1).
Diamond-Blackfan is a macrocytic anemia, and it does produce physical abnormalities in the described systems (1).
## [1] 8
The cell lines most associated with the positively expressed gene set in CCLE Proteomics 2020 were KASUMI1 (haematopoeitic and lymphoid tissue), SW403 (large intestine), U937 (haematopoeitic and lymphoid tissue), HT55 (large intestine), and a further two large intestine cell lines.
KASMU1 is an acute myeloid leukemia line (1).
SW403 is a colon adenocarcinoma (2).
U937 is acute monocytic leukemia (3), but was at some point contaminated with K-562 which is a chronic myeloid leukemia (4).
HT55 is colon carcinoma (5).
1 https://web.expasy.org/cellosaurus/CVCL_0589 2 https://web.expasy.org/cellosaurus/CVCL_0545
3 https://web.expasy.org/cellosaurus/CVCL_0007
4 https://web.expasy.org/cellosaurus/CVCL_0004 5 https://web.expasy.org/cellosaurus/CVCL_1294
These suggest a relationship to blood cells.
MOLT-4 is a lymphoblastic leukemia cell line (immature lymphocytes) (1,2), and promyelocytic leukemia involves WBC promyelocytes (3).
CD34+ stem cells lead to the hematopoeitic cells in the body, but is that the CD34+ cell they are describing? (4)
Burkitts Lymphoma is related to abnormal B lymphocytes (5), and as has been discussed it is related to the MYC gene.
1 https://www.sigmaaldrich.com/catalog/product/sigma/cb_85011413?lang=en®ion=SE
2 https://www.cancer.gov/types/leukemia/patient/child-all-treatment-pdq
3 https://rarediseases.org/rare-diseases/acute-promyelocytic-leukemia/
4 https://cellero.com/blog/ask-a-scientist-progenitor-vs-cd34-cells/
5 https://www.cancerresearchuk.org/about-cancer/non-hodgkin-lymphoma/types/burkitt-lymphoma
The cell types related to the positive expression continue in a similar way in the ProteomicsDB 2020, with the highest combined score cells being two variants of hematopoietic (K-562), followed by bone (U2-OS) variants. These cell lines are chronic myeloid leukemia (1) and osteosarcoma (2), respectively.
1 https://web.expasy.org/cellosaurus/CVCL_0004
2 https://web.expasy.org/cellosaurus/CVCL_0042
The ARCHS4 tissues tells a slightly different story, with bulk kidney, human embryo being higher. The human embyro tissue is perhaps due to the CD34+ cells and their differentiation. The adjusted p values here as a whole however are not very low.
In the top ARCHS4 the top two cell lines are both colon adenocarcinoma cell lines (1,2).
1 https://web.expasy.org/cellosaurus/CVCL_0320
2 https://web.expasy.org/cellosaurus/CVCL_0218
In ESCAPE there is some return to the relationship with MYC, as well as H3K36me3, which is a histone mark related to DNA repair (1).
1 https://cellandbioscience.biomedcentral.com/articles/10.1186/s13578-020-0374-z
The NCI-60 Cancer Cell Lines database similar supports the COLO205 is related to the positively correlated gene set, and the other colon adenocarcinoma cell line HT29 is not far below. HCC2998 is also a colon adenocarcinoma cell line (1). HEPG2 is a hepatoblastoma cell line (2), and HEK 293T is a fetal kidney cell (3). We do also see MOLT4 again, the acute lymphoplastic leukemia cell line from earlier (4).
1 https://web.expasy.org/cellosaurus/CVCL_1266
2 https://web.expasy.org/cellosaurus/CVCL_0027
3 https://web.expasy.org/cellosaurus/CVCL_0063
4 https://web.expasy.org/cellosaurus/CVCL_0013
## [1] 1
The misc category on enrichr provided some interesting results in that there might be some association between testis seminoma and Rloops, but as the sample numbers are sequential (622-624) they could just be from the same individual who may be overexpressing these genes for other reasons. The sample numbers are not in the same sequence when referring back to the GSE8607 in GEO (1). However, testicular germ cell tumors were associated with UCK2 kinase, which had a high combined score for this gene set.
Given the amount of ribosome-related results this could be masking the information about the R-loops from other sources.
Some repeated/high combined score genes/gene product results: MYC + VEGFA, EKLF, POU5F1, E2Fs, XRN2, KAT2A, CNBP?
Otherwise, the areas involved can again be separated into groups: 1. Growth
growth + differentiation (in various systems)/stem cells + angiogenesis
2. Blood cells
hematopoiesis/erythropoiesis/lymphopoiesis + RBCs
Blood disorders
anemia(mostly macrocytic, Diamond-Blackfan) + pure red cell aplasia acute/chronic myeloid leukemia + lymphoblastic leukemia + promyelocytic leukemia + Burkitts Lymphoma
Regulation of cell functions
tumor suppression/apoptosis + DNA damage response + DNA repair + cell-cyle regulation (G2-M) + Nonsense-mediated decay
Retinal cancer
retinal cancer/retinoblastoma
Carcinomas
colorectal cancer/colon adenocarcinoma/carcinoma + lung cancer/lung adenocarcinoma
Sarcomas
Ewing sarcoma + osteosarcoma
CNS and cancers/disorders
CNS tissues + Limb hypertonia + Meningoencephalitis + glioblastoma + astrocytoma
Kidney tissue
Kidney/fetal kidney tissues
Additional cancer types
testicular cancer/testis seminoma + hepatoblastoma
Additional cell functions
Protein targeting + viral gene expression + inhibiting ubiquitin-protein transferase
“ENCODE_and_ChEA_Consensus_TFs_from_ChIP-X” “ARCHS4_TFs_Coexp” “TF_Perturbations_Followed_by_Expression” “TRRUST_Transcription_Factors_2019” “lncHUB_lncRNA_Co-Expression” “ChEA_2016” “BioPlanet_2019” “WikiPathways_2019_Human” “KEGG_2019_Human” “ARCHS4_Kinases_Coexp” “MSigDB_Hallmark_2020” “Elsevier_Pathway_Collection” “GO_Biological_Process_2018” “GO_Molecular_Function_2018” “GO_Cellular_Component_2018” “Human_Phenotype_Ontology” “COVID-19_Related_Gene_Sets” “DisGeNET” “DSigDB” “ARCHS4_IDG_Coexp” “Rare_Diseases_GeneRIF_ARCHS4_Predictions” “Rare_Diseases_AutoRIF_ARCHS4_Predictions” “CCLE_Proteomics_2020” “Human_Gene_Atlas” “ProteomicsDB_2020” “ARCHS4_Tissues” “ARCHS4_Cell-lines” “ESCAPE” “NCI-60_Cancer_Cell_Lines” “Enrichr_Submissions_TF-Gene_Coocurrence” “Reactome_2016” “BioPlex_2017” “Jensen_DISEASES” “Jensen_COMPARTMENTS” “ClinVar_2019” “DepMap_WG_CRISPR_Screens_Sanger_CellLines_2019” “Rare_Diseases_AutoRIF_Gene_Lists” “Virus_Perturbations_from_GEO_down” “Virus_Perturbations_from_GEO_up” “Allen_Brain_Atlas_10x_scRNA_2021” “NCI-60_Cancer_Cell_Lines” “Disease_Perturbations_from_GEO_up”