Find the biology behind neurodegeneration.
BioSymetrics combines multimodal patient data, machine learning, and experimental validation to define meaningful disease subgroups and advance biomarkers, targets, and therapeutic hypotheses. Our Parkinson’s disease (PD) and Alzheimer’s disease (AD) programs are two core applications of the same approach.
One diagnosis can contain several disease trajectories.
Neurodegenerative diseases can present with early signs and subclinical biological signals that are difficult to distinguish from normal variation or connect to later progression. Clinical records, imaging, molecular measurements, and experimental data can each capture part of the picture, making it harder to identify which patient groups are likely to progress differently or respond differently to treatment.
We bring these signals together so partners can move from broad diagnostic labels to patient groups with clearer biological and development relevance.
From patient patterns to testable hypotheses.
We integrate longitudinal clinical and molecular data to identify phenotypic subgroups and progression patterns. We then connect those patterns to biomarkers, pathways, and targets, prioritizing the hypotheses that merit experimental follow-up.
Our approach weaves a common thread across disease areas. We start with patient biology, define what differs between subgroups, and test the most useful hypotheses in relevant systems.
Turn patient heterogeneity into development decisions.
In PD, our analysis of more than 650 participants in the Parkinson’s Progression Markers Initiative, roughly 4,500 proteomic probes, and multiple years of follow-up identified three outcome-linked subtypes. They were associated with fast motor progression, combined cognitive and neurological decline, and broad functional impairment.
These findings provide progression-linked biomarkers, pathways, and targets for precision diagnostic and therapeutic development. We use the subgroups to guide program design, including patient selection, trial enrichment, progression monitoring, treatment-response assessment, target prioritization, and partner-facing decision packages.
With our partner BrainStorm Therapeutics, we are combining clinically defined PD subtypes with patient-derived human midbrain organoids and complementary in vivo systems. Our collaboration will test subtype-associated mechanisms, biomarkers, and targets, with an initial data readout targeted for the first half of 2027. It connects patient-level patterns with biological evidence and development decisions.
Bring more of the disease signal into view.
In AD, we apply the same workflow to combine imaging and molecular signals. In benchmarks on the Alzheimer's Disease Neuroimaging Initiative dataset, models that combined MRI, genomic, and metabolomic data improved diagnostic accuracy by up to 35% compared with MRI-only models. We also ran more than 22,000 MRI processing runs, showing how distributed analysis and automated parameter optimization can support heterogeneous data at scale.
With our partner Moleculera Biosciences, we are developing an inflammatory biomarker panel that combines neuroinflammatory biomarkers with personal and family history measures and our analytics. The panel supports earlier disease indicators and trajectory monitoring alongside imaging and molecular data.
Could patient subtyping inform your neurodegeneration program?
If you have a PD or AD dataset, candidate biomarker, target, or therapeutic hypothesis, we would welcome a conversation about the next evidence-generating step, from patient stratification and biomarker work to experimental validation.
Partner with us