Intro
HASU is an independent research initiative working on structure-aware biomedical AI.
The data biology produces is rarely shapeless. A genome carries a history; a patient record unfolds in time; the scientific literature argues with itself. We study how machine learning can represent biological, clinical, and scientific data in forms closer to the processes that generate them, rather than flattening everything into the same strings, images, and tables. That spans evolutionary histories, genomic variation, clinical trajectories, wearable signals, multimodal health records, and the structure of scientific evidence.
We also write. Alongside the technical work, HASU publishes shorter essays and notes on ideas worth exploring, research philosophy, scientific infrastructure, publication systems, and the broader question of how biomedical AI should be built, evaluated, trusted, and used.
People
Ao Zhang
PhD researcher working across biomedical machine learning, genomics, digital phenotyping, and scientific computing.
Yusuf Abdulle
PhD researcher working across multimodal biomedical AI, clinical data, genetics, and disease progression modelling.
Contact
For general correspondence, email [email protected].
Disclaimer and acknowledgements
HASU is an independently run research and editorial initiative. Essays, notes, commentary, and independently developed technical work published here represent the views of the individual authors unless otherwise stated. They do not represent the views, position, or official research output of any university, funder, employer, commercial sponsor, or collaborating organisation.
Some of the authors’ research is carried out as part of funded doctoral work and institutional research programmes. Any output that arises from, substantially relates to, or uses resources from such work is governed by the relevant institutional, funding, collaboration, intellectual-property, data-governance, and publication-review requirements. Where such work is discussed or linked from this site, it is identified and acknowledged appropriately and is not published ahead of any required institutional review.
HASU does not publish confidential information, sponsor-confidential material, controlled-access data, clinical data, participant data, or other personal data arising from the authors’ institutional research.
Content on HASU is provided for research discussion and general informational purposes only. It is not legal, medical, financial, or investment advice, and does not constitute a commercial offer.