Drop in a patient's full medication list. xrayGraphDB searches 149,361 known interactions across 1,933 drugs in under 2 milliseconds — mapping every enzyme conflict, every cascade risk, and telling you which single change helps the most.
A patient on 10 medications has 45 possible drug pairs to check. Current tools check them one at a time. Nobody checks all 45. Nobody checks the cascades.
Drug interactions aren't just pairs. They're chains. When Drug A inhibits the enzyme that metabolizes Drug B, Drug B's blood levels rise, and Drug B now interacts differently with Drug C. This is a graph traversal problem.
Amiodarone inhibits CYP3A4 → Simvastatin can't metabolize → blood levels rise 4x → rhabdomyolysis risk. No pairwise table catches this chain.
4 drugs competing for CYP3A4. The enzyme is overloaded. Which drug is the bottleneck? Which can move to a different metabolic pathway?
"Removing Drug X eliminates 7 of your 12 interactions and reduces polypharmacy risk by 58%." One click to see the impact of every possible removal.
Warfarin, digoxin, lithium, phenytoin — drugs where small level changes cause toxicity. Any enzyme interaction with these is flagged CRITICAL.
"Replace Simvastatin with Rosuvastatin: same therapeutic class, zero interactions with your current list." Graph query: same condition, no edges.
All interactions, cascades, enzyme loads, removal impacts, and alternatives computed in under 1 millisecond. Add or remove a drug — the entire graph updates instantly.
The analysis runs inside xrayGraphDB as native graph traversals over 149,361 interaction edges and 14 metabolic enzyme pathways — not external rules engines or lookup tables.
Drug → Enzyme → Drug → Enzyme → Drug. Variable-length path queries find cascade chains that no pairwise database can detect.
Composite risk scoring across multiple dimensions: interaction severity, enzyme load, narrow therapeutic index, polypharmacy count.
Remove one drug from the graph. Instantly compute the difference in interaction count, risk score, and enzyme load. GRAPH_DIFF in microseconds.
148 of the most commonly prescribed drugs include hand-curated CYP450 enzyme substrate/inhibitor/inducer data, enabling cascade detection and enzyme load analysis. The remaining 1,785 drugs have interaction pairs and severity from the DDInter 2.0 database. Enzyme families tracked: CYP3A4, CYP2D6, CYP2C9, CYP2C19, CYP1A2, CYP2B6, CYP2E1, CYP2C8, CYP2J2, UGT1A3, UGT1A4, UGT1A9, UGT2B7, and MAO-A.
This is a technology demonstration, NOT a clinical tool. Do not use for actual medical decisions. Always consult a qualified healthcare provider or pharmacist. Drug interaction data sourced from DDInter 2.0 open-access database and hand-curated CYP450 literature — not a complete clinical database.
The Live Layers SPA exposes 14 multi-hop signal layers (PRR, symptom clusters, recall blast radius, pharmacogenomic cascade, hidden twins…). The Polypharmacy Analyzer is the classic patient-drug-list workflow — autocomplete from 186K drugs, interaction matrix from real FAERS co-reports, shared side-effect overlap, class-redundancy, removal-impact deltas.
Launch Live Layers Launch Polypharmacy Analyzer