Key Reflections from AD/PD 2026

Written by Daniel McCartney PhD, Principal Data Scientist – Translational Genomics

This year’s AD/PD brought together more than 5,500 delegates from over 70 countries, reinforcing its status as one of the world’s leading meetings in neurodegenerative disease research. A clear message across the conference was that the field is moving toward greater biological precision, with genetics, AI, and mechanism-driven target discovery increasingly shaping both research strategy and therapeutic development.

A recurring theme was that neurodegenerative diseases can no longer be treated as single, uniform entities. Instead, the science is moving toward subtypes, staging, co-pathology, and molecularly defined mechanisms. Progress was framed as a move beyond broad diagnostic labels and toward biologically informed models that can support more precise interventions. Bart De Strooper, for example, argued for understanding Alzheimer’s disease through sequential biological phases and inflection points rather than through the late clinical endpoint of dementia alone.

Disease heterogeneity, biomarker-based stratification, and combination therapies were recurring themes. Integrative frameworks that connect molecular mechanisms, biomarkers, and clinical features are increasingly required to capture progression, interaction, and heterogeneity. Supporting hypothesis generation across stages and co-pathologies helps move therapeutic thinking upstream and aligns target development strategies more closely with underlying biology.

Genetics provided a strong example of this transition. The focus is no longer simply on identifying risk loci, but on using genetics to uncover mechanism. Danielle Posthuma emphasised the value of linking genetic architecture to pathways, vulnerable cell types, heritability, and potential drug targets. Aviv Regev extended that idea by showing how genetics can be integrated with single-cell and spatial atlases to identify disease-relevant programmes in specific cell states and tissues.

Significant value lies in translating genetic signal into actionable biological insights. Integrating genetics with cellular atlases and functional data, rather than viewing them in isolation, enables the identification of convergent mechanisms, key pathways, and context-specific disease risk. Grounding genetic associations in relevant cell types and disease states strengthens target prioritisation and informs decisions around timing and mechanism of intervention.

AI emerged as another major theme, framed less as future promise and more as a practical necessity. Regev also highlighted AI-enabled cellular atlases, causal modelling, and iterative “lab-in-the-loop” systems that can connect genetics, pathology, cell states, and clinical traits in ways that support target prioritisation and medicine development. More broadly, speakers showed how AI is already beginning to shape trial design, biomarker integration, assessment tools, and patient selection.

The impact of AI is greatest when it is embedded within R&D workflows, rather than treated as a standalone capability. By transforming complex, multi-modal datasets into interpretable models, AI can connect genetics, biological mechanisms, biomarkers, and clinical outcomes in a way that directly supports scientific decision-making. This supports more continuous learning across clinical development and allows therapeutic hypotheses and strategies to be refined as evidence accumulates.

In practice, these shifts are increasingly reflected in how discovery approaches are evolving. Integrating genomic, clinical, and real-world data within unified analytical workflows is becoming essential to move from fragmented evidence to actionable insight, enabling faster and more confident decision-making in early-stage drug discovery.

In Parkinson’s disease, we have applied this approach by integrating large-scale genetic datasets with transcriptomic, proteomic, and other multi-omics data to build a connected view of disease biology. Using the analytical tools in DiscoveryX, this enabled the identification of disease-associated genes, proteins, and pathways within specific tissues and cell types, while linking these insights to broader cardiometabolic and neurodegenerative traits. By embedding these relationships within our developing causal, knowledge-driven framework GraphX, we have been able to strengthen target validation, prioritise biologically grounded mechanisms, and significantly reduce time to insight, supporting faster and more confident early-stage decision-making.

Frameworks such as DiscoveryX, alongside analytical tools like PleioGraph, are designed to embed genetic evidence directly into discovery workflows. By connecting multi-omics data with clinical context, they support the identification of causal relationships, reduce false positives, and strengthen target validation. This ultimately helps R&D teams prioritise more effectively and progress therapies with greater confidence.

The overall takeaway from AD/PD 2026 is a field moving toward greater integration, sharper biological definition, and more precise therapeutic thinking. AI and genetics are now central to how mechanisms are understood, targets are prioritised, and future therapies may be developed. As these capabilities continue to mature, their value will increasingly be realised through their integration into scalable, decision-focused discovery frameworks that can translate complex data into meaningful scientific and clinical progress.