Amir Hossein Amanzadi

Doctoral Research Fellow @ University of Oslo

I'm just a human being. The descriptions below are social statuses I've collected along the way, which you may or may not find interesting.

My name is Amir, a researcher with an R&D mindset, working across industry and academia, from hands-on wet lab research to large-scale AI deployment. One goal has stayed constant: building methods and tools that help improve human health and longevity. I'm currently a Doctoral Research Fellow in Medicine at the Institute of Clinical Medicine, University of Oslo, and Oslo University Hospital, working at the Centre for Precision Psychiatry on multi-omics and agentic AI for psychiatric and neurological disease.

I hold a Master's in Pharmaceutical Sciences from Uppsala University and double majored in Chemistry and Mechanical Engineering at Sharif University of Technology, with roles at Karolinska Institutet, Centre for Molecular Medicine, and Celeris Therapeutics. On the AI side, I've worked on AI solution architecture, graph neural networks, and generative AI at Prepaire Labs, Algorithmic Dynamics Lab, and my own ventures, Zetazi, Cognition Research, Shenasa, and Yar AI.

More in my CV. The interesting parts are harder to list, so feel free to email me.

Research Perspective

Human biology is dynamic. Disease does not emerge from a single molecule or a single moment. It emerges from interactions across scales, from atoms to populations, unfolding over time.
Most of our medical data are cross-sectional. Genetics, imaging, and multi-omics each give a precise but static slice of a moving system. To understand disease, we need to connect these snapshots into something that evolves the way the body does.
I suspect this is a representation problem before it is a data problem. Biology is relational, so it needs an architecture that captures relationships and lets them change. Graph-based knowledge and agentic AI with long-term memory seem like reasonable places to start.
The obstacle is scale. Biology runs at the molecular, cellular, organ, and population level at once, and no single tool spans all of them. I work at that integration layer, combining multi-omics, genetics, and imaging with graph neural networks and agentic AI, trying to build models that hold up in both space and time.
The distant dream is a digital twin, a real-time biological runtime of human physiology across all scales. We are nowhere near it, and that absence is part of why medicine moves slowly. It may not arrive in my lifetime, but it seems worth working toward. This is why I work with large population cohorts and national medical registries at Oslo University Hospital, where the data depth makes the direction possible.

Work

Publication image for Protein–protein interaction prediction for targeted protein degradation

Protein–protein interaction prediction for targeted protein degradation 2022

Oliver Orasch, Noah Weber, Michael Müller, Amir Amanzadi, Chiara Gasbarri, Christopher Trummer

International Journal of Molecular Sciences

project paper supplement

Expand bibtex

@InProceedings{IJMS2022,
  author     = {Oliver Orasch and Noah Weber and Michael Müller and Amir Amanzadi and Chiara Gasbarri and Christopher Trummer},
  title      = {Protein–protein interaction prediction for targeted protein degradation},
  booktitle  = {International Journal of Molecular Sciences},
  year       = {2022},
}

Publication image for Predicting safe drug combinations with Graph Neural Networks (GNN)

Predicting safe drug combinations with Graph Neural Networks (GNN) 2021

Amir Amanzadi

Uppsala University

project paper code

Expand bibtex

@InProceedings{UU2021,
  author     = {Amir Amanzadi},
  title      = {Predicting safe drug combinations with Graph Neural Networks (GNN)},
  booktitle  = {Uppsala University},
  year       = {2021},
}

Publication image for Explainable polypharmacy side effect prediction with Siamese graph convolutional neural networks

Explainable polypharmacy side effect prediction with Siamese graph convolutional neural networks 2021

Amir Amanzadi, Narsis Kiani

4th RSC-BMCS Conference, Royal Society of Chemistry Lightning Presentation

project video code

Expand bibtex

@InProceedings{RSC2021,
  author     = {Amir Amanzadi and Narsis Kiani},
  title      = {Explainable polypharmacy side effect prediction with Siamese graph convolutional neural networks},
  booktitle  = {4th RSC-BMCS Conference, Royal Society of Chemistry},
  year       = {2021},
}

Publication image for Both Tough and Soft Double Network Hydrogel Nanocomposite Based on O-Carboxymethyl Chitosan/Poly(vinyl alcohol) and Graphene Oxide: A Promising Alternative for Tissue Engineering

Both Tough and Soft Double Network Hydrogel Nanocomposite Based on O-Carboxymethyl Chitosan/Poly(vinyl alcohol) and Graphene Oxide: A Promising Alternative for Tissue Engineering 2020

Ali Pourjavadi, Zahra Tehrani, Hamid Salami, Farzad Seidi, Anahita Motamedi, Amir Amanzadi, Ehsan Zayerzadeh, Meisam Shabanian

Polymer Engineering & Science

project paper

Expand bibtex

@InProceedings{PES2020,
  author     = {Ali Pourjavadi and Zahra Tehrani and Hamid Salami and Farzad Seidi and Anahita Motamedi and Amir Amanzadi and Ehsan Zayerzadeh and Meisam Shabanian},
  title      = {Both Tough and Soft Double Network Hydrogel Nanocomposite Based on O-Carboxymethyl Chitosan/Poly(vinyl alcohol) and Graphene Oxide: A Promising Alternative for Tissue Engineering},
  booktitle  = {Polymer Engineering & Science},
  year       = {2020},
}

Publication image for Designing a new multifunctional peptide for metal chelation and Aβ inhibition

Designing a new multifunctional peptide for metal chelation and Aβ inhibition 2018

Amir Shamloo, Mohsen Asadbegia, Vahid Khandan, Amir Amanzadi

Archives of Biochemistry and Biophysics

project paper

Expand bibtex

@InProceedings{ABb2018,
  author     = {Amir Shamloo and Mohsen Asadbegia and Vahid Khandan and Amir Amanzadi},
  title      = {Designing a new multifunctional peptide for metal chelation and Aβ inhibition},
  booktitle  = {Archives of Biochemistry and Biophysics},
  year       = {2018},
}

Talk image for Application of Graph Neural Networks (GNN) in Drug Discovery

Application of Graph Neural Networks (GNN) in Drug Discovery 2022

Uppsala University, Sweden

slides recording