Predictive Systems Biology & Advisory
Moving Beyond Descriptive Catalogs
Bridging high-dimensional single-cell, spatial omics, and causal network modeling into reproducible, perturbation-predictive biology.
Dr. Qingzhou Zhang
Computational Systems Biologist & Bioinformatician
The Core Scientific Thesis
“Modern omics produces enormous amounts of descriptive data. The next challenge is not simply identifying what differs between conditions, but understanding biological states, predicting how systems change, and determining which interventions can alter those trajectories.”
The Problem: High-Throughput Biology Is Trapped in Cataloging
For over 15 years, my work has focused on extracting causal biological mechanisms from high-dimensional genomics data—from mapping human mitochondrial interactomes and transcriptional repression complexes to resolving cell-fate transitions in complex disease models.
Throughout this trajectory, a systemic limitation became clear: modern omics excels at descriptive catalogs, but struggles with predictive causation. We generate massive cell atlases, cluster cell types, and produce exhaustive tables of differentially expressed genes.
Yet identifying that gene X is upregulated twofold in condition B versus A rarely explains why the transition happened, how the network will respond to a novel perturbagen, or which combination of interventions can steer a pathological trajectory back to health.
I founded Omics With Johnson as an independent practice to bridge this gap: translating high-dimensional observations into predictive perturbation models, deterministic pipelines, and publication-ready computational insights.
Three Pillars of the Practice
To maintain scientific rigor, intellectual independence, and deterministic execution, my work is organized around three distinct pillars:
Mechanistic Systems Biology
Investigating causal transcriptional control, mitochondrial protein communities, and dynamic cellular states. Co-authored publications in Nature Communications, Molecular Cell, Cell Reports, and Advanced Functional Materials.
Fixed-Scope Consulting
Direct scientific advisory for academic labs, biotech teams, and core facilities. Providing study design reviews, advanced perturbation modeling, and manuscript-critical computational analysis through defined, milestone-driven scopes.
Deterministic Open-Source Tooling
Developing open-source packages (Bioconductor SMAD, PyPI scfeatureprofiler), spatial processing pipelines, and modular reproducible workflows (Bash/Conda Epi-Flow, container and Nextflow-ready) designed for bit-for-bit computational reproducibility.
Operating Principles
- Causal Reasoning Over Observational Correlation: Prioritizing experimental designs, counterfactuals, and perturbation modeling that test causal mechanisms rather than passive marker clustering.
- Deterministic Reproducibility: Every analysis runs within version-locked containers and pipeline orchestrators. If a finding cannot be re-executed from raw counts or reads, it is an unverified hypothesis.
- Stress-Testing AI & Foundation Models: Refusing to accept deep representation embeddings without rigorous benchmarking against strong linear baselines and out-of-distribution biological splits.
- Milestone-Driven Deliverables: Consulting and collaboration are structured around clear scopes, reproducible code repositories, and publication-ready analytical artifacts.