Patient-led knowledge system ยท early stage

MOLI

Medical Observation through Linked Insights. A structured, longitudinal, computable model for complex and rare medical histories โ€” built from one unusually complex patient case, designed for many.

Concept phase โ€” seeking scientific & conceptual sparring partners

Fragmented by design, not by accident

Complex, multisystem, rare-variant patient histories generate information across many specialties, institutions and document formats. Individual findings are repeatedly assessed in isolation, while their temporal, anatomical and potentially biological relationships remain difficult to examine โ€” even though the underlying data already exists.

MOLI began with a concrete case: pathogenic or clinically relevant variants in SDHA and FH, combined with a long history of multifocal bone and organ lesions, endocrine abnormalities, neurological symptoms, imaging findings, pathology, interventions and shifting diagnostic interpretations. The goal is to transform such fragmented records into something inspectable โ€” without collapsing their complexity or uncertainty.

A knowledge graph, not a chronology

Every clinical element receives a stable identity and remains traceable to its original source. The model connects:

01
Genetic variants & pathwaysVariant-level findings linked to associated biological mechanisms.
02
Phenotypic findings & symptomsPatient-observed and clinically documented, kept distinguishable.
03
Anatomy: region, structure, lateralityHierarchical โ€” e.g. pelvis โ†’ ilium โ†’ right โ€” not flat tags.
04
Imaging & pathology observationsLinked to the structure and timepoint they describe.
05
Diagnoses & diagnostic uncertaintyConfidence is a first-class field, not an afterthought.
06
Treatments, procedures & observed responseWhat was done, and what changed afterward.
07
Temporal development & clinical eventsThe case as a timeline, not a snapshot.
08
Source documents & provenanceEvery element traceable back to the letter or report it came from.
09
Hypothesised relationshipsKept clearly, structurally separate from verified clinical fact.

Compatible, not reinvented

We are examining how MOLI could align with established frameworks while retaining information conventional records often lose โ€” particularly chronology, uncertainty, treatment response and patient-observed relationships.

Human Phenotype Ontology MONDO GA4GH Phenopackets FHIR Standard biomedical terminologies

What we'd value a second opinion on

Which entities and relationships are essential for a scientifically meaningful longitudinal rare-disease graph?

How should verified findings, clinical interpretations and patient-generated hypotheses be distinguished without discarding potentially relevant signals?

How can such a system remain interoperable with established standards while preserving the complexity of an individual patient's history?

At what point could an individual patient model become useful for cohort comparison or research hypothesis generation?

Founder & patient lead

MOLI began with one unusually complex patient history. Its larger purpose is to help make other medically "invisible" patterns examinable, shareable and scientifically useful.

Huanita Mijatovic-Pflanz
Founder & Patient Lead, MOLI

Not asking for formal supervision, medical advice, or a substantial time commitment โ€” an occasional scientific and conceptual sparring partner during this early phase.

Get in touch

If this concept is relevant to your work, we'd be glad to send a concise visual overview and arrange an informal 20โ€“30 minute conversation.

contact@moli-health.org