1. Homepage
The homepage of konnect2prot is the starting point. Click the "Click here to start" button to navigate to the application page.
Figure 1: The k2p homepage.
2. Dashboard
Figure 2 shows the K2P dashboard. This will be the first window that pops up when you enter the K2P portal.The flow chart gives an overview of the portal's workings. There will be four tabs: Data, EDA (Exploratory Data Analysis), Visualization, and Network Analysis. The data upload will be done through the Data tab.
Figure 2: The k2p Dashboard.
3. Data Tab
1. In the data tab, there will be an upload sidebar where you can upload your file either by dragging it
or by clicking the Drag and Drop files here tab, as shown in Figure 3.
2. The file will automatically open and show the data preview at the right side of upload tab.
3. A preliminary analysis of the uploaded data will be conducted.
4. The results will be shown in graphical forms, including a Box Plot, Mean-Variance Trend, Density
plot, and QQ plot, as shown in Figures 4 and 5.
Figure 3: Upload your file in the Data Tab sidebar.
Figure 4: Uploaded Data Preview.
Figure 5: Different Plots of uploaded Data
4. EDA
1. As shown in fig 6 there is a pre-processing Parameters side tab where you can select Normalization method, Scaling Method, Alpha Value (Significance Threshold).
2. The normalization method includes log2, log10, and None.
3. The scaling method includes Min-Max, Standard Scaling or None.
4. Alpha value as per your experiment needs.
Figure 6: Pre-processing Parameters.
5. Group Naming includes an input box for defining the names of groups in a binary fashion. You have
to select only for Set 1, and Set 2 will be automatically selected. For example, the data can be
grouped into CASE and CONTROL, as shown in Figure 7.
6. After clicking on Submit the DEGs Table will be generated alongside with Summary of Sample and DEG as shown in Figure 8(a)
Figure 7: Group Naming.
Figure 8: DEGs and Summary.
7.User can also explore 'Pathway Explorarion' tab to see desired pathway containing DEGs, 'Complex Exploration' tab to see desired complex containing DEGs, 'Similarity Exploration' tab to explore the expression similarity of selected gene in the selected group , 'Expression Distribution'(Upto 5 cross validation) tab to see the expression profile of the selected gene(s) in both groups by box plot representation. As shown in Figure 8(b)
Figure 8b: Exploration and Expression Distribution.
5. Visualization
1. In this section, there will be a Volcano plot, pathway enrichment, PCA of complete Gene Expressions, and
Significant Gene Expression, as shown in Figures 9 (a) and 9 (b).
2. In the volcano plot, the slider is used to change the Log2 fold change value.
Figure 9(a): Volcano plot and Pathway enrichment.
Figure 9(b): PCA plot.
4. Users can upload a DEG file in the specified format (as shown in Figure 10) to visualize the volcano plot and perform pathway enrichment analysis.
5. However, for generating PCA plots, a complete gene expression file is required.
6.Once the analysis is complete, users can proceed to the Network tab for further exploration.
Figure 10: Direct DEG Upload from Visualization Sidebar
6. Network Analysis
Here, we have searched k2p using an example gene, "CDK1". The protein-protein interaction (PPI) network of CDK1 and its first neighbours will be constructed at the right-hand side panel. This network can further be filtered using "localization", "molecular functions", "biological processes", "tissue-specificity" or "pathways". An example is shown in Figure 11. For a smooth visualisation, k2p provides different layout options, which can be found in the layout tab, as shown in Figure 11. Click the analysis button to find the enriched pathways and ontologies, multi-disease interactome, and topological analysis.
Figure 12: Information on PDB structures, ligands, interaction type and disease mutation.
You may find out how many PDB structures are available for a protein by clicking on it in the created PPI network. The ligand panel provides information on small compounds and their mode of action for the query protein. The "mutation in disease" panel contains information on disease-specific mutations of this protein (if any). The information about the mode of interaction can be accessed by clicking an edge in the network. An example is shown in Figure 12.
a. Enrichment panel
By clicking the "analysis" button, the enrichment panel will display the enriched pathways and processes for the proteins in the constructed PPI network. Additionally, k2p also provides the protein class abundance and the multi-disease landscape of the proteins in the PPI network. This information is shown in Figure 13.
Figure 13: The Enrichment panel displays information on enriched pathways, processes, the multi-disease landscape, and protein class abundance
b. Topological panel
The topological panel (see Figure 14) illustrates the results of three critical measures of centrality: degree, betweenness, and closeness centrality. A plot of degree versus betweenness plot is also included to identify the proteins that act as hubs and bottlenecks in the constructed PPI network. For a detailed understanding of the different centrality measures and their application please refer to [1].
c. Spreaders
We have identified the influential spreaders in the network and augmented it with other auxiliary information. Identifying a set of influential spreaders in complex networks plays a crucial role in effective information spreading, which is identified using the voterank algorithm [2]. In this approach, all nodes vote in a spreader in each turn, and the voting ability of neighbours of the elected spreader will be decreased in subsequent turns. The identified triggers could be explored during the investigation for various applications, such as potential drug targets. As illustrated in Figure 15, k2p identifies the triggers in the PPI network and files their topological properties, cellular localisation, class, available PDB complexes and ligands. Afterwards, a clustergram of pathways Figure 8 related to the spreaders is shown to give an idea of which pathways are modified by the network's top spreaders. This cluster gram and the high tissue specificity of these influential spreaders can be exported in .png format. The spreaders can be targets or triggers, depending on the context of the study.
Figure 14: The topological panel. You can use the designated buttons to search for a protein, navigate through the results, and even sort the analysis results.
Figure 15: Result panel for the influential spreaders.
d. Spreader-Hallmark associations
Protein-Hallmark associations are another crucial property of k2p. Every disease is driven by specific characteristics or hallmarks. Identifying proteins associated with the hallmarks helps identify new therapeutic targets with more specific pharmacological activity. Various drugs are deliberately developed for specific molecular targets that involve these hallmarks [3]. Addressing this, k2p incorporates two crucial aspects of drug discovery: protein-hallmark associations and protein-signalling pathway associations Figure 17. The latter will enable the identification of not just intra-pathway deregulation but also the interdependence of pathways. Again, this information can be utilized to deduce the pleiotropic effects of a large number of genes on distinct pathways that contribute to the development of specific disease characteristics or traits.
Figure 16: Clustergram of the pathways.
Figure 17: Association of spreaders with A) disease hallmarks B) signalling pathways
References
- 1 . Minoo Ashtiani, Ali Salehzadeh-Yazdi, Zahra Razaghi-Moghadam, Holger Hennig, Olaf Wolkenhauer, Mehdi Mirzaie, and Mohieddin Jafari. A systematic survey of centrality measures for protein-protein interaction networks. BMC Systems Biology, 12(1):1_17, 2018.
- 2 Jian-Xiong Zhang, Duan-Bing Chen, Qiang Dong, and Zhi-Dan Zhao. Identifying a set of influential spreaders in complex networks. Scientific Reports, 6:27823, 2016.
- 3 Natalia Bailón-Moscoso, Juan Carlos Romero-Benavides, and Patricia Ostrosky-Wegman. Development of anticancer drugs based on the hallmarks of tumor cells. Tumor Biology, 35(5):3981_3995, 2014.6