Skip to main content

​
The Problem You Are Trying to Solve

“I have a set of lead compounds, and I want to understand both their direct on-target effects and their indirect off-target effects, including how they engage broader biological pathways and cellular programs.”
Understand On Target And Off Target Biological Pathway Engagement For Lead Compounds
At the lead optimization stage, it is no longer sufficient to know that a compound binds a target. You need to understand:
  • Whether the compound robustly engages the intended target in a biologically meaningful way
  • Which secondary proteins or pathways may be perturbed downstream
  • Whether observed phenotypes arise from on-target mechanism, off-target liabilities, or network-level effects
  • How molecular binding events translate into cellular state changes
This workflow bridges molecular-scale binding analysis with systems-level biological response modeling, helping teams move from “binds the target” to “behaves as expected in biology.”

​
Solution

This workflow integrates binding chemistry engines with transcriptomic pathway analysis to connect molecular engagement with biological consequence. The primary analysis chain is: On-Target Binding Validation → Off-Target Binding Survey → Transcriptomic Response Analysis → Pathway & Network Interpretation. This allows users to:
  • Confirm intended target engagement
  • Identify plausible off-target interactions
  • Observe how cells respond to compound treatment
  • Resolve whether effects are direct, indirect, or compensatory
Other engines (MD, FEP, quantum chemistry, generative chemistry) can be integrated as follow-ups once mechanisms are identified.

​
What Data Do I Need to Provide?

Required
  • Lead compound structures (SMILES or CSV)
  • Primary target protein structure or sequence
  • Cellular transcriptomic data (e.g., scRNA-seq) from treated vs control conditions
Recommended
  • Time-resolved transcriptomic data (multiple doses or timepoints)
  • Known pathway annotations for the disease context
Optional
  • Off-target protein panels
  • Comparative compounds (tool compounds, inactive controls)

​
Workflow

  1. Validate Direct On-Target Engagement
Start by confirming that leads engage the intended target in a plausible and specific manner.
Using Docking and (if applicable) Boltz or BoltzGen Co-Folding, users:
  • Evaluate binding poses and interaction patterns
  • Confirm engagement of known functional residues
  • Compare multiple leads for consistency of on-target binding
This step establishes a mechanistic anchor for downstream interpretation, giving a validated on-target binding hypotheses, and key interaction residues and motifs.
  1. Survey Potential Off-Target Binding
Next, assess whether leads may engage additional proteins that could drive indirect effects. Using Virtual Screening and Docking against a curated off-target panel, users can:
  • Identify secondary binding candidates
  • Flag proteins involved in signaling, metabolism, or stress responses
  • Distinguish selective compounds from promiscuous binders
This step generates candidate off-target hypotheses, not conclusions, in the form of a ranked list of plausible off-target interactions per lead.
  1. Analyze Cellular Response with scRNA-seq
Now connect molecular interactions to biological outcomes. Using scRNA-seq Analysis, users can:
  • Perform differential gene expression (DEG) analysis between treated and control cells
  • Identify transcriptional programs altered by compound exposure
  • Resolve heterogeneity across cell populations
This step answers:
  • What changes when the compound is applied?
  • Which cell states are most affected?
This results in differential expression profiles across cell types and conditions.
  1. Map Effects to Pathways and Gene Regulatory Networks
Translate gene-level changes into mechanistic understanding. Using Automated Target ID and Gene Regulatory Network Analysis, users can:
  • Identify upstream regulators driving observed transcriptional changes
  • Map altered genes to known signaling pathways
  • Predict cascade effects downstream of target engagement
This helps distinguish:
  • Direct on-target pathway modulation
  • Indirect off-target or compensatory responses
This step gives a pathway-level interpretation of compound effects with ranked regulators and network hubs.
  1. Model Temporal and Dose-Dependent Effects (Optional)
If time-series data is available, deepen the analysis. Using Temporal Omics Analysis, users can:
  • Model how gene expression evolves after compound exposure
  • Identify early vs late response programs
  • Separate primary effects from downstream adaptation
This step is especially valuable for:
  • Chronic dosing scenarios
  • Pathway rewiring and resistance studies
This produces dynamic models of pathway engagement over time.
  1. Integrate and Interpret with Revilico Agent
Finally, synthesize results across molecular and cellular layers. Using Revilico Agent, users can:
  • Ask mechanistic “why” questions across datasets
  • Compare on-target binding hypotheses with observed pathway changes
  • Generate testable biological hypotheses for follow-up experiments
This gives cohesive mechanistic narratives linking chemistry to biology.

​
Results

  • Clear differentiation between on-target and off-target effects
  • Mechanistic insight into pathway-level engagement
  • Reduced ambiguity in phenotypic interpretation
  • Stronger confidence in lead progression or redesign decisions

​
Now What? I understand how my leads engage biology, but what’s next?

Common next steps include:
  • Refining chemistry to amplify desired pathways and suppress undesired ones
  • Designing focused validation experiments
  • Prioritizing leads with clean, interpretable mechanisms
  • Integrating findings into toxicity or efficacy optimization workflows

​
Why Revilico?

Revilico uniquely connects:
  • Molecular binding engines for direct interaction analysis
  • Single-cell transcriptomics for biological response resolution
  • Network and temporal modeling for pathway-level insight
  • AI-assisted reasoning to unify complex datasets
This enables teams to move beyond single-target thinking and design molecules with predictable, system-aware biological behavior.