Clinical sequence modelling

Projects rooted in clinical event sequences, electronic health records, representation learning, and translational phenotype discovery.

Translational research thread

Phenotype fingerprinting

A line of work using clinical sequence models to identify rare, under-recognised, or hard-to-characterise disease patterns.

The aim is to use latent clinical trajectories to surface patients or cohorts whose records contain early, partial, or indirect evidence of a phenotype before conventional coding makes it obvious. This work is intended to support earlier recognition, better cohort definition, and more testable translational hypotheses.

Clinical sequence modellingRare-case and under-recognised phenotype discovery
Research programme

ORCA

A programme built around the idea that electronic health records can be treated as a language, with implications for representation, translation, and generation.

The aim is to move beyond treating EHRs as flat tables of variables and instead study the statistical structure, translation behaviour, and generative possibilities of clinical event sequences. ORCA asks whether coded records can be read, aligned across systems, translated into clinical language, and used to reason about plausible futures.

Working paper in progressLanguage-of-EHR framing
Programme-level thread

ASCEND research arc

A broader research arc spanning ASCENDgpt, FlatASCEND, and ORCA, focused on modelling structured clinical trajectories.

The aim is to develop a coherent sequence of EHR foundation-model work: from phenotype-aware representation learning, through autoregressive clinical trajectory generation, to interpretation and language-like translation. The arc is designed to make clinical sequences modelable, inspectable, and ultimately more useful for real clinical research questions.

ASCENDgptFlatASCENDORCA

Scientific discovery systems

Tools for navigating scientific possibility space: finding transferable ideas, mapping research routes, and keeping literature-scale evidence usable.

Knowledge exploration system

NEXUS

A system for identifying where scientific ideas, methods, and representations might transfer across fields.

The aim is to move beyond ad hoc analogy-making. NEXUS uses large-scale scientific literature, historical examples of successful cross-domain transfer, and structured novelty checks to identify where a method or concept from one field might become useful in another. It began with clinical and biomedical examples, but the broader ambition is more general: computational support for navigating scientific possibility space.

AI-assisted discoveryCross-domain transferLiterature-scale evidenceHypothesis generation
Experimental route-analysis toolkit

Physarum

A local toolkit that turns a concise research starting position into route cards, safe exploratory checks, review packets, route memory, and a next-step memo.

The aim is to make uncertain research frontiers easier to inspect without pretending to automate scientific judgement. Physarum extracts candidate routes from a safe starting-position file, runs route-aware filaments for literature, terminology, analogy, skeptic, and implementation pressure, gathers public-metadata and safe-local signals into reviewable packets, preserves pursue / park / reject decisions, and writes an everyday decision artifact: physarum-return.md.

Route-aware filamentsReview packetsRoute decisionsphysarum-return.mdSafe public-metadata checks
Domain literature intelligence

glucose.ai

A literature intelligence and discovery surface for diabetes and AI / machine learning work.

The aim is to make the diabetes and AI literature easier to monitor, search, and turn into useful research leads. glucose.ai acts as a public-facing discovery surface for papers, themes, and emerging methods in diabetes, metabolic medicine, and clinical AI.

Public websiteDiabetes + AI / ML discoveryLiterature radar

Research infrastructure

Local systems and workflow patterns for making complex research programmes easier to run, review, and resume.

Research operating system

pi / Cog workflows

Research and coding infrastructure for organising sources, tasks, synthesis, experiments, and project state in a more usable system.

The aim is to keep complex research work connected enough that ideas can develop over time rather than being lost. Current patterns include outcome loops for rubric-based artifact evaluation, Nightshift routines for daily review and continuity, guarded site-maintenance checks, and agent-supported workflows for turning messy context into durable research memory.

Outcome loopsNightshift reviewCogX continuityAgent-supported research workflows