Steven Bergner — Computational Research

Mathematical computing, scientific data, and reproducible research systems

I am a computational researcher in Greater Vancouver, Canada, with a long research history at Simon Fraser University. My work combines mathematical computing, scientific data analysis, and reproducible research systems. I hold a PhD in Computing Science from Simon Fraser University, where my earlier work addressed multidimensional parameter spaces, visualization, sampling, and experimental design.

Contact and record: GitHub · ORCID · Google Scholar · doctoral thesis at SFU ·

Research programme

Scientific computing and experimental design

My research lineage includes sampling and refinement in high-dimensional spaces, visual and interactive methods for understanding complex computational systems, and exact or replayable methods for comparing scientific alternatives.

Computational biology and perturbation systems

I study how perturbation-response programmes can be reconstructed, compared, and transported across cell types, biological contexts, and experimental platforms. The immediate work uses public single-cell perturbation data and emphasizes exact dataset identity, prospective controls, reliability limits, and informative refusal when a carrier cannot support the claim being tested.

One current programme asks when perturbation-response structure remains recognizable across neuronal and microglial models relevant to neurodegeneration. This is computational methods research: it is not medical advice, a clinical model, or a treatment claim.

Evidence-governed computation

I build research software that keeps proposals, evidence, permissions, refusals, and realized outputs distinct. The practical aim is not merely to produce a result, but to make clear what the result depends on, where it applies, what failed, and how an independent reader could reproduce or challenge it.

Visual and data-intensive computing

My earlier work includes image analysis, volume rendering, deformable models, visual parameter-space exploration, and data-intensive research infrastructure. These remain useful foundations for current work on biological populations and scientific decision systems.

Open collaboration questions

I welcome focused exchanges with researchers who can help establish an exact, lawful route from a published study to a reusable computational carrier. Current high-value questions include:

My preferred starting point is a small, concrete identity or provenance question—not a broad partnership commitment. See the collaboration note for the information that makes a dataset scientifically actionable.

Research practice

I separate exploratory findings from confirmation, freeze important comparisons before opening outcome values, preserve negative results, and avoid extending conclusions beyond the jurisdiction of the data. Public descriptions deliberately omit private implementation details while keeping claims, sources, and limitations inspectable.

Selected research history

About this site

This is a concise public research profile. It links established work and describes current directions without treating work in progress as validated evidence. The site is built with Jekyll and hosted on GitHub Pages; its source is public.