After processing and normalization of the datasets, the following 3 pairwise comparisons were performed:
- PTB (GSE18809) and Type 1 Diabetes (GSE29142)
- PTB (GSE18809) and Obesity (GSE110551)
- PTB (GSE18809) and Periodontitis (GSE33774)
3.1 Pairwise comparison of PTB with Type 1 diabetes
3.1.1 Identification of overlapping DEGs
Under the threshold of p-value < 0.05, a total of 555 overlapping DEGs were found out through the comparison of datasets of PTB and type 1 diabetes. Among these, 289 DEGs were down regulated and 266 were upregulated.
3.1.2 Functional enrichment analysis
Cutoff criteria of p-value < 0.05 and counts > 10 was specified for functional enrichment analysis of DEGs. GO terms enrichment analysis showed that each of the 3 categories such as Biological processes (BP), Cellular components (CC) & Molecular function (MF) were enriched with 19, 16 & 12 terms respectively. BPs which were mainly found enriched with DEGs included ‘apoptotic process’, ‘signal transduction’, ‘cell-cell adhesion’, ‘regulation of cell cycle’, ‘protein phosphorylation’. Significantly enriched CCs were ‘nucleus’, ‘nucleoplasm’, ‘nucleolus’, ‘cytosol’ and ‘’cytoplasm’. While in the MF group, the DEGs were mainly enriched in ‘protein binding’, ‘actin binding’, ‘enzyme binding’, ‘ATP binding’ and ‘protein kinase activity’. Fig. 1 shows the results of the GO terms enrichment analysis. KEGG pathways enrichment analysis resulted in 17 significant pathways including ‘hsa04151: PI3K-Akt signaling’, ‘hsa04015: Rap1 signaling’, ‘hsa04068: FoxO signaling’, ‘hsa04062: Chemokine signaling’ and ‘hsa04630: Jak-STAT signaling’ pathways. The results of the KEGG pathways enrichment analysis are shown in Fig. S1.
3.1.3 PPI network and identification of hub genes
Protein-protein interaction (PPI) network constructed under the confidence score of ≥ 0.4 which resulted in 449 nodes and 1293 edges (see Fig. S2). Top 10 hub genes from the PPI network were retrieved on the basis of the highest degree through cytoHubba. Among these hub genes, PTPRC, VEGFA, TLR4, MAPK1, MAPK14 and APP genes were upregulated while LYN, PTEN, CCND1 and MDM2 were downregulated (see Table S2).
3.1.4 Modules extraction and their KEGG pathway enrichment analysis
Three significant modules were found in the whole PPI network by following the criteria of score > 4 and nodes > 5 as shown in Fig. 2 (a, b, c). Score of the 1st module was 8 which contained 18 nodes and 68 edges, while the score of 2nd module was 6.444 having 19 nodes and 58 edges, and 3rd module was found with score of 4.857 consisting of 29 nodes and 68 edges. Pathways analysis of the modules by following the criteria of p-value < 0.05 showed that ‘ubiquitin mediated proteolysis pathway’ are enriched in module 1 (Fig, 2d), while 3 other pathways ‘Inflammatory bowel disease (IBD)’, ‘Malaria’ and ‘Proteoglycans in cancer’ are found significant in the module 2 (Fig. 2e). In module 3, no significant pathway was found enriched under the specified criteria.
3.2 Pairwise comparison of PTB with obesity
3.2.1 Identification of DEGS
Under this comparison, a total of 595 overlapping DEGs were found. Among these, 293 genes were down regulated and 302 genes were upregulated.
3.2.2 Functional enrichment analysis
Threshold criteria of p-value <0.05 and counts > 10 was specified for functional enrichment analysis of the DEGs. GO terms enrichment analysis showed that each of the 3 categories of GO terms BP, CC & MF were enriched with 25, 15 & 19 terms respectively. Examples of 5 BPs which were mainly found enriched included ‘in utero embryonic development’, ‘MAPK cascade’, ‘ubiquitin-dependent protein catabolic process’, ‘protein polyubiquitination’ and ‘intracellular signal transduction’. Significantly enriched 5 examples of CCs were ‘nucleus’, ‘nucleoplasm’, ‘perinuclear region of cytoplasm’, ‘cytosol’ and ‘cytoplasm’. While in the MF group, ‘DNA binding’, ‘Poly (A) RNA binding’, ‘nucleotide binding’, ‘chromatin binding’ and ‘ubiquitin protein ligase binding’ are the 5 examples. Fig. 3 shows the results of the GO terms enrichment analysis. KEGG pathways enrichment analysis resulted in 13 significant pathways including ‘hsa04068:FoxO signaling’, ‘hsa04120:Ubiquitin mediated proteolysis’, ‘hsa04014:Ras signaling’, ‘hsa04010:MAPK signaling’ and ‘hsa04068:FoxO signaling’ pathways. The results of the KEGG pathways enrichment analysis are shown in Fig. S3.
3.2.3 PPI network and identification of hub genes
Protein-protein interaction (PPI) network constructed under the confidence score of ≥ 0.4 resulted in 456 genes with 1701 edges as shown in Fig. S4. Top 10 hub genes from the PPI network were retrived on the basis of the highest degree through cytoHubba. Among these hub genes, three genes (UBE2D2, BCL2L1, CTNNB1) were found down regulated while 7 genes (FOXO3, CDC42, HSPA4, CDH1, EGFR, MAPK1, ACTG1) were found upregulated (see Table S3).
3.2.4 Modules extraction and their KEGG pathway enrichment analysis
Four significant modules were found in the whole PPI network by following the criteria of score > 4 and nodes > 5. Module 1 had 16 nodes with a score of 16 (Fig. 4a), module 2 was found with a score of 8.083 and 25 nodes (Fig. 4b), module 3 had 9 nodes and score of 4.5 (Fig. 4c) and 4th module was found with score of 4.435 and 24 nodes (Fig. 4d). Pathways analysis of the modules following the criteria of p-value < 0.05 depicted that three clusters were enriched with the pathways under the cutoff criteria of p-value < 0.05. Pathway analysis of cluster 2 did not return any results under prescribed criteria. In cluster 1, ubiquitin mediated proteolysis pathway was enriched with 6 genes (Fig. 4e); in cluster 3, Lysine degradation pathway was found with 3 genes (Fig. 4f) and cluster 4 showed 3 significant pathways, ‘Bacterial invasion of epithelial cells’ with 3 genes, ‘adherens junction pathway’ with 5 genes and ‘endometrial cancer pathway’ with 3 genes (Fig. 4g).
3.3 Pairwise comparison of PTB with periodontitis
3.3.1 Identification of DEGS
Under this comparison, a total of 687 overlapping DEGs were found. Among these, 358 genes were down regulated and 329 genes were upregulated.
3.3.2 Functional enrichment analysis
Threshold criteria of p-value < 0.05 and counts > 10 were specified for functional enrichment analysis. GO terms enrichment analysis showed that each of the 3 categories BP, CC & MF were enriched with 10,14 & 10 terms respectively. BPs which were mainly found enriched with DEGs included ‘macroautophagy’, ‘cell-cell adhesion’, ‘ubiquitin-dependent protein catabolic process’, ‘positive regulation of gene expression’ and ‘heart development’. Significantly enriched CCs were ‘cytoplasm’, ‘cytosol’, ‘cytoskeleton’, ‘nuclear membrane’ and ‘nuclear envelope’. While in the MF group, the DEGs were mainly enriched in ‘transcription factor binding’, ‘protein kinase activity’, ‘protein binding’, ‘protein kinase activity’ and ‘Poly (A) RNA binding’. Fig. 5 shows the results of the GO terms enrichment analysis. KEGG pathways enrichment analysis resulted in 7 significant pathways including ‘hsa04120:Ubiquitin mediated proteolysis’, ‘hsa04015:Rap1 signaling’, ‘hsa04810:Regulation of actin cytoskeleton’ and ‘hsa04510:Focal adhesion’ and ‘hsa04068:FoxO signaling’ pathways. The results of the KEGG pathways enrichment analysis are shown in Fig. S5.
3.3.3 PPI network and identification of hub genes
PPI network constructed under the confidence score ≥ 0.4 resulted in 586 nodes and 1958 edges as shown in Fig. S6. Top 10 hub genes from the PPI network were retrived on the basis of the highest degree through cytoHubba. Among these hub genes, one hub gene (KEAP1) was found down regulated while 9 hub genes (BPTF, EZR, HUWE1, KRAS, NCBP2, PIK3R1, PPP2R1A, PTGS2, and VCL) were found upregulated (see table S4).
3.3.4 Modules extraction and their KEGG pathway enrichment analysis
Four significant modules were found in the whole PPI network by following the criteria of score > 4 and nodes > 5. Module 1 had 14 nodes with score of 14 (Fig. 6a), module 2 was found with a score of 9.333 and 13 nodes (Fig. 6b), module 3 had 11 nodes and score of 8.8 (Fig. 6c) and 4th module was found with score of 4.151 and 54 nodes (Fig. 6d). Pathways analysis of the modules following the criteria of p-value < 0.05 depicted that three clusters were enriched. Pathway analysis of cluster 2 did not return any results under prescribed criteria. In cluster 1, ‘ubiquitin mediated proteolysis’ pathway was enriched with 6 genes (Fig. 6e); in cluster 3, ‘longevity regulating’ pathway with 3 genes, ‘autophagy’ with 6 genes and ‘spinocerebellar ataxia’ with 5 genes were found enriched (Fig. 6f) while cluster 4 showed 3 significantly enriched pathways including ‘RNA degradation’ with 4 genes, ‘long term depression’ with 3 genes and ‘citrate cycle’ with 3 genes (Fig. 6g).
3.4 Analysis of DEGs through DIA
The top 20 KEGG pathways found significant in the PTB dataset are shown in the Fig. 7. These results were generated through DIA, an excel based software. The results show the top 5 categories, their subcategories and respective biological pathways ranked by the “Impact” values. The column labelled with ‘Impact’ shows the magnitude of each pathway, larger the bar the greater the impact is. The column labelled ‘Flux’ shows the direction of the impact. The color ranges from green (inhibited) to red (induced) for each of the KEGG pathway.