Polypharmacological Target Prediction

The Polypharmacological Target Prediction tool builds a protein interaction network from the genes shared across the diseases (or genes) you select, and identifies potential multi-target drug candidates and their known drugs from DGIdb. Choose a Disease entry point (select multiple diseases and find the genes, pathways, and drug targets common to all of them) or a Gene entry point to start directly from a set of genes, then pick a dataset — Curated or Curated + Text Mining.

Results span 7 tabs, in order:

GeDiPNet Polypharmacological Target Prediction Common Genes tab: 681 common genes for Diabetes mellitus type 2 AND Alzheimer disease — 1. Common Genes — every gene shared across your selection (681 here)

1. Common Genes lists every gene shared across all the diseases (or genes) you selected — 681 for this pair — sortable, searchable, and exportable.

Enriched Pathways lists the KEGG pathways your common genes are significantly concentrated in — those with a Benjamini–Hochberg adjusted p-value (q-value) below 0.05 — showing each pathway's matched genes, hypergeometric p-value, fold enrichment and q-value (see "How to read the statistics" under Enrichment Analysis). This pathway list is also what the Network Visualization, Network Analysis and Summary tabs are built from. In Version 1, a Show previous-method list switch displays the list as this tab showed it under the previous method, for tracing earlier results only.

GeDiPNet Polypharmacological Target Prediction Network Visualization tab: gene-pathway interaction network — 2. Network Visualization — the gene-pathway interaction network

2. Network Visualization draws the gene-pathway interaction network directly, with Fit to Screen, Zoom, Reset, and Export PNG controls.

GeDiPNet Polypharmacological Target Prediction Network Analysis tab: hub and bottleneck genes with centrality scores — 3. Network Analysis — every gene scored by degree, closeness, and betweenness centrality

3. Network Analysis scores every gene in the network by weighted degree, closeness centrality, and betweenness centrality, and flags Hub genes (highly connected nodes) and Bottleneck genes (key connectors between modules) — filterable via the All Genes / Hub Genes / Bottleneck Genes buttons above the table.

GeDiPNet Polypharmacological Target Prediction Summary tab: 102 critical targets with 764 known drugs, top target LINGO1 — 4. Summary — the ranked shortlist of critical targets and their known drugs

4. Summary narrows the network down to the critical polypharmacological targets — 102 for this pair, with 764 associated known drugs — each row showing the target gene, its associated pathways, a known drug, the interaction type, and whether it's a hub or bottleneck node.

GeDiPNet Polypharmacological Target Prediction Drug Targets tab: drugs with known and predicted targets — 5. Drug Targets — every candidate drug, with an "Analyze further" button into the Drug Deep-Dive

5. Drug Targets flips the view around: one row per drug (751 analyzed here), showing its known targets, its predicted targets from this network, and an "Analyze further" button that opens that specific drug in the Drug Deep-Dive tab.

GeDiPNet Polypharmacological Target Prediction Drug Deep-Dive tab for a specific drug and its primary target — 6. Drug Deep-Dive — one drug, studied in depth across 5 on-demand panels

6. Drug Deep-Dive studies one drug/target pair in depth across 5 panels, each loading only when you click its own "Analyze" button (so the page doesn't slow down loading all 5 up front): Clinical Trial Status (real ClinicalTrials.gov results), Tissue Expression (GTEx data for the target gene), Transcriptomic Cross-reference (Expression Atlas and, for cancers, TCGA/cBioPortal data), Side Effects (on-label SIDER data plus off-label OFFSIDES signals), and ADMET Screening (absorption/distribution/metabolism/excretion/toxicity predictions, including blood-brain barrier penetration).

Help

Polypharmacological Target Prediction