Biocellion brings modeling skills and high-performance software—our customers bring the biological domain expertise. We work alongside each customer to encode what they know and measure as a computational model. Utilizing that model to run a simulation maps their knowledge to a prediction. Any discrepancies between that prediction and their observations lead to corrections and subsequent improvements—as the process iterates the model becomes increasingly predictive.

Simultaneously, scientists hone their understanding of the relative importance of biological mechanisms in determining outcomes.

Two examples of such partnerships are those with Celgene and with Procter and Gamble. Both companies presented posters introducing the collaborative work at the 10th Anniversary of the Interagency Modeling and Analysis Group's Multiscale Modeling Consortium in 2017 on the NIH campus.

Celgene uses Biocellion to model therapies using the company's pharmaceutical products for multiple myeloma, a blood cancer resident in bone marrow. In the image, over a dozen cell types and a comparable number of molecular species (not visible) participate in the dynamics of the bone microenvironment. The large cancerous cells inhibit maintenance of the bone surface leading to the pitting visible, for example, in the lower right quadrant. The impact of various combinations of drugs each having known mechanisms of action on slowing or eliminating these destructive forces can be studied using variations of this model.


Procter and Gamble models skin growth and response to materials in consumer care products such as bandages and sunblock. Growing the outermost layers of skin in a laboratory petri dish takes about one month. When Biocellion is used to model these experiments, the simulation of the month of growth takes just hours on powerful computers. Qualitative comparisons of morphology and quantitative comparisons of cell population by type match up, providing reassurance the model is on track to augment if not replace physical experiments.

Screen Shot 2017-12-20 at 10.51.02 AM.png

Biovision's virtual histology feature enables interactive exploration within simulation output with greater ease than physical histology. In this example, molecular concentrations in the environment may be explored as well: points of equal humidity are represented here by isosurfaces. Dead cells at the surface of the skin are shown shedding.

Grid 014.png

In both partnerships, members of the Biocellion team worked closely with scientists at the respective companies to encode domain knowledge as a Biocellion model customized to the client's proprietary scenarios.

A typical engagement is described below as a sequence of developmental phases over the course of 3-4 years. You can also download a PDF of the table below here.

Click to enlarge