Virtual Cell Simulator for Biomolecular Condensate (BMC) Cellular Entry
This repository contains a lightweight Virtual Cell simulator for studying how biomolecular condensates (BMCs) enter cells, with an emphasis on early membrane interaction, entry pathways, and membrane order remodeling.
The model is designed as a mechanism-driven, hypothesis-testing scaffold, rather than a fully resolved biophysical simulation. It explicitly encodes experimentally testable rules derived from live-cell imaging, GP (C-Laurdan), CRISPR perturbations, and membrane biology.
Scientific Motivation
Biomolecular condensates span a wide size regime (nano- to micron-scale) and engage distinct cellular entry mechanisms. However, these mechanisms are often conflated under generic “uptake” labels.
This virtual cell framework is built to disentangle:
Nano-BMC entry
ATP-dependent
FLOT1 / CLTC1 (clathrin-mediated) endocytosis
Preserves membrane order (higher GP)
Micro-BMC entry
ATP-independent or weakly ATP-coupled
Lipid raft + actin “landing”
TMEM16F activation → lipid scrambling
Strong membrane order disruption (GP drop)
Elevated membrane tearing risk
Shared post-entry sorting
RAB5A (early endosome)
SNX33 (membrane remodeling / sorting)
The simulator formalizes these hypotheses into a dynamic, perturbable in silico cell.
What This Model Is (and Is Not)
This model IS:
A virtual cell logic engine
Mechanism-encoded (not purely data-driven)
Designed for CRISPR KO / chemical perturbation reasoning
Directly mappable to imaging and GP readouts
This model is NOT:
A full molecular dynamics simulation
A quantitative predictor without calibration
A replacement for experiments
Think of it as a computational figure + hypothesis validator.
Model Overview Core State Variables
contact_on(t) – BMC–membrane engagement
site_protein – local enrichment at contact site
active_TMEM16F(t) – lipid scrambling activity
vesicle(t) – endocytic vesicle progression
deform_load(t) – membrane mechanical stress
tear_risk(t) – membrane rupture proxy (micro-BMC)
GP(t) – membrane order (C-Laurdan proxy)
Key Proteins Encoded Category Proteins Lipid raft / landing RFTN2 Cytoskeleton ACTA1 Lipid scrambling TMEM16F Endocytosis FLOT1, CLTC1 Endosomal sorting RAB5A, SNX33 Installation python >= 3.9 pip install numpy matplotlib
No other dependencies required.
Running the Simulator
- Micro-BMC Entry python virtual_cell_bmc_entry.py --mode micro --minutes 5 --plot
Expected behavior:
Strong GP decrease
High TMEM16F activation
Elevated membrane deformation
Possible tear risk
- Nano-BMC Entry python virtual_cell_bmc_entry.py --mode nano --minutes 5 --plot
Expected behavior:
Gradual vesicle formation
Limited GP disruption
Strong CLTC1/FLOT1 recruitment
- CRISPR Knockout (in silico) python virtual_cell_bmc_entry.py --mode nano --ko CLTC1 --plot
python virtual_cell_bmc_entry.py --mode micro --ko TMEM16F --plot
Use this to:
Predict uptake failure modes
Design CRISPR screens
Interpret low-uptake cell sorting
- Cholesterol Depletion (e.g. MβCD) python virtual_cell_bmc_entry.py --mode micro --cholesterol_factor 0.5 --plot
Directly maps to GP and raft-dependent phenotypes.
Outputs Console Summary
Internalization time
Minimum / final GP
Vesicle progression
Tear risk (micro-BMC only)
Plots (optional)
GP vs time
Vesicle & deformation dynamics
Protein recruitment kinetics
CSV (optional) --out_csv results/micro_BMC.csv
CSV is Prism- and Python-ready.
How to Calibrate to Experiments
This model is intentionally parameter-light.
You can calibrate it using:
Live-cell recruitment kinetics (TIRF / spinning disk)
C-Laurdan GP time series
CRISPR KO phenotypes
Uptake timing distributions (flow or imaging)
Typical workflow:
Fit recruit_k to imaging slopes
Fit GP coefficients to C-Laurdan data
Validate KO phenotypes qualitatively
Use model for mechanism discrimination, not absolute prediction
Intended Use Cases
Designing CRISPR uptake screens
Rationalizing nano vs micro BMC behavior
Framing mechanism figures for high-impact papers
Supporting claims of non-endocytic entry
Building toward a Virtual Cell / digital twin framework