Lundbeck needed a faster, more efficient way to identify and qualify drug targets for migraine and other headache disorders because the biology is complex, there are thousands of possible genetic links and pathways to assess, and traditional literature review plus laboratory validation is time-consuming and costly.
Reported impact
Lundbeck used the AI approach to find potential drug targets for headache disorders faster, support existing research programs with structured evidence, identify known and novel gene-headache associations, and deploy a framework now used across its research organization.
Lundbeck worked with Accenture to build an AI-driven in silico discovery workflow centered on a domain-specific knowledge graph for headaches and migraines. The implementation ingests Accenture’s large-scale clinical dataset, including 54 million electronic medical records, together with public data on biological pathways, genetics, and diseases. These sources are structured into a semantic knowledge graph that links genes, diseases, symptoms, drugs, pathways, and other biological and clinical concepts so researchers can analyze relationships across datasets. Accenture then applied machine learning models on top of the graph to predict probable gene-disease links from the available evidence and infer new associations not explicitly stated in source data. The system also provides explanations by surfacing the key graph facts behind each prediction, allowing Lundbeck researchers to generate pathophysiological hypotheses, prioritize promising targets, and support ongoing research programs with structured evidence.
Clinical data from 54 million electronic medical records and public biological data sources are ingested into Accenture’s large-scale dataset and organized in a domain-specific headache and migraine knowledge graph. Machine learning models analyze graph relationships to predict gene-disease links, and the resulting ranked targets plus explanation signals are delivered to Lundbeck researchers for hypothesis generation and prioritization.
"Accenture then applied machine learning models to predict the probability of gene-disease links in the knowledge graph, based on the available evidence."
"Lundbeck decided to use a combination of knowledge graphs, a way to identify and analyze the link between different data sets, and AI to augment its research process and generate hypotheses in silico"
"Their dataset consists of 54 million electronic medical records that include information on clinical characteristics and public data sources including biological pathways, genetics, and diseases."
"Together, they built a domain-specific knowledge graph for headaches and migraines, using Accenture’s large-scale clinical dataset."
"Using AI, Lundbeck was able to find potential drug targets associated with headache disorders, such as migraine, 80% faster than traditional methods."
"developed a new approach to interrogating biological datasets that are now being widely used in Lundbeck’s Research departments"
"the Accenture team also provided explanations for their predictions, showing Lundbeck the key facts in the knowledge graph that influenced the decisions."
"the team demonstrated that the approach could identify 77% of known gene-headache associations and suggested novel biological targets for drug discovery."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
The Organization JSON-LD lists addressLocality "Valby", addressCountry "DK", and numberOfEmployees value 5617.
Related
In-house Sample Jul 28, 2026 Novartis Novartis R&D needs to turn large volumes of scientific, clinical, operational, and real-world data into faster, higher-quality decisions across drug discovery and development, reducing cycle times, surfacing safety… less than 10 percent of drug candidates historically reach the market AI models · AI-enhanced analytics · Generative AI In-house Sample Apr 13, 2026 Elevance Health Elevance Health needed to find fraud, waste, and abuse in very large volumes of claims, payment, contract, and medical-record data so it could detect billing errors, suspicious coding patterns, and improper payments… across an annualized volume of nearly 23,000 reviews In-house Sample Feb 24, 2026 Sanofi Sanofi needed a faster, more objective way to decide which R&D programs to advance, accelerate, or stop while weighing costs, clinical trial enrollment timelines, unmet need, competition, and each program's probability… more than one billion data points across Sanofi Aily Labs · Plai 