Skip to main content
The Problem You are Trying to Solve:
“I want to understand how my disease of interest works biologically (what pathways, mechanisms, and cellular programs are driving it) before committing to target selection or downstream drug discovery workflows.”
Hero Image
Before pursuing structure prediction, molecular docking, or de novo compound design, it is critical to understand the biological context of a disease. This includes identifying dysregulated pathways, cell-type-specific transcriptional programs, and temporal dynamics that underlie disease onset and progression. However, extracting these insights from high-dimensional omics data is often nontrivial and requires specialized bioinformatics expertise, custom pipelines, and significant interpretation effort. A centralized, interpretable workflow that connects differential expression, target prioritization, and temporal biology enables researchers to move from raw data to mechanistic understanding with confidence. Solution
Revilico provides a unified biological discovery layer designed to help users understand disease mechanisms and pathways at the cellular level. Using transcriptomics-driven target identification workflows, the platform allows users to interrogate how a disease operates biologically before advancing to structural biology or chemistry-driven pipelines. This disease-understanding layer can serve as a preceding step to downstream target validation and structure-based discovery, enabling users to ensure that their hypotheses are grounded in biological signal rather than isolated assumptions. In addition to data-driven analyses, users may optionally leverage RevilicoGPT as a contextual intelligence layer to synthesize known biology, pathway annotations, and mechanistic hypotheses alongside their experimental results. What Data Do I Need to Provide?
  • Control and Experimental h5ad files
  • Bring some understanding of your target disease of interest
Workflow
  1. Differential Gene Expression (DGE) → Identify What is Changing
The first step in understanding a disease is determining what molecular programs are altered relative to a control or baseline state. Users upload control and experimental scRNA-seq datasets (h5ad files). Revilico computes differential gene expression across relevant cell populations, identifying genes that are significantly upregulated or downregulated in the disease or experimental condition. To interpret the results:
  • Upregulated genes often indicate activated pathways, stress responses, or compensatory mechanisms
  • Downregulated genes may reflect loss of normal function, differentiation defects, or pathway suppression
  • Patterns across specific cell types help distinguish cell-intrinsic drivers from systemic effects
At this stage, the goal is not to select a final target, but to form an initial hypothesis about which biological processes are disrupted.
  1. Automated Target Identification → Prioritize What Matters
Once transcriptional changes are identified, the next challenge is determining which genes are most biologically and therapeutically relevant. Revilico’s Automated Target ID layer ranks candidate genes emerging from DGE by integrating:
  • Magnitude and consistency of expression change
  • Cell-type specificity and relevance
  • Signal robustness across samples or conditions
This step reduces hundreds or thousands of DE genes into a shortlist of biologically meaningful targets. To interpret the results:
  • High-ranking targets are likely to be drivers rather than passengers
  • Cell-type-restricted targets may suggest precision intervention opportunities
  • Broadly dysregulated targets may indicate core disease machinery
Rather than forcing a single “best” answer, this step provides a ranked landscape of plausible intervention points.
  1. Temporal Omics Analysis → Understand How the Disease Evolves
Many diseases are dynamic systems. Understanding when biological changes occur is as important as knowing what changes. For datasets with temporal structure (e.g., disease progression, treatment timepoints), Revilico models gene expression as a function of time. This reveals dynamic patterns in pathway activation and target behavior. To interpret the results:
  • Early-changing genes may represent initiators or causal drivers
  • Late-changing genes may reflect downstream effects or compensatory responses
  • Transient expression peaks can indicate regulatory switches or checkpoints
Temporal insights help differentiate targets that are merely associated with disease from those that may be strategically actionable depending on intervention timing. Results
  • Key dysregulated genes and pathways associated with the disease
  • Prioritized target candidates grounded in transcriptomic evidence
  • Cell-type-specific and temporal insights into disease mechanisms
  • A biologically informed hypothesis of how the disease operates at the molecular level.
These results collectively provide a mechanistic understanding of the disease, rather than a single-point target guess. RevilicoGPT can be used alongside these analyses to contextualize identified genes with known pathways and disease biology, summarize mechanistic hypotheses supported by both literature and data, and assist in interpreting complex multi-gene or temporal patterns. This added layer complements experimental data without replacing it, helping users reason about why observed patterns may be occurring. Now what? Now that you have a strong understanding of your disease biology and candidate targets:
  • If you want to validate or explore the target structure → proceed to AlphaFold or OpenFold within our platform.
  • If you have compounds or libraries → advance to binding chemistry or quantum chemistry workflows.
  • If you do not yet have compounds → use de novo library generation to design molecules for your selected target, or make use of our on hand liquid stock or powder libraries to be delivered after structure based drug design has been concluded.
Why Revilico?
This Revilico workflow enables users to query the availability of experimental data before proceeding to AI structure prediction. Utilizing these transcriptomic based workflows allows you to understand the deeper mechanics driving your disease of interest to investigate or validate therapeutic avenues and proteins to target.