Explore the cell space for options in biomanufacturing.
Choose the right cell line, spot the right clone and find the perfect environmental conditions for maximum production for cell products like mAbs, biomaterials and more.

MoNA tackles the challenge of accurately mapping and digitally representing the intricate relationships between multi-omic phenotypic profiles, transomic dynamics, and the physical-chemical conditions that govern biological processes.
Genomic, transcriptomic and proteomic information obtained from wet-lab assays.
System biology based ontologic relationships are extracted from all available literature.
Culture media, temperature, gases, pH and other cellular environments conditions are included.
A structured pipeline to create cell to cell, genes to genes, genes to transcripts, cell identity to overexpressed genes or even experiment to experiment.
Biological systems are truly complex, multilayered and diverse, which makes them especially challenging to represent in a digital form.
We started by creating the Big Omic Database, unifying genetic, transcriptomic, and proteomic data into a machine-learning-ready framework.
To add structure and meaning, we built a Knowledge Graph, layering scientific insights that capture not just correlations, but causal relationships. By combining data-driven approaches with human insight, we transformed raw data into a rich, meaningful representation of cellular behavior.
Using a network-based perspective, we define cell identity through hierarchical relationships and causal impact, rather than solely observing the presence or concentration of biomolecules or environmental conditions.
Focusing on network architecture shows how small changes in critical biomolecules can trigger bigger systemic effects than large changes in non key cell elements.
We compressed each cell state to its most essential form using autoencoders, a deep learning technique. By compressing cell states through a narrow bottleneck and reconstructing them, we achieved high signal-to-noise ratios, creating minimal yet accurate representations of each cell’s identity.
Having compressed and represented each of our cells into a latent space means we can extract them from that multidimensional representation to whatever map we find useful. The closer two cell identities are, the more their behavior and phenotypes will be alike in the real world. This system can also work in reverse, using a generative approach to simulate and explore potential cell states, expanding our understanding of cellular behavior.
Choose the right cell line, spot the right clone and find the perfect environmental conditions for maximum production for cell products like mAbs, biomaterials and more.
Increase or decrease RNA expression or protein abundance, or even change physicochemical conditions and understand how they will affect specific cell identities.
Find the best phenotype for a given therapy, but also the best starting cell types and specific protocols to generate your desired cell product.
Wet-lab assay design and decision-making are costly, labor-intensive, and inherently uncertain due to the complexity of biomolecular systems. Use MoNA to search for the most informative wet-lab assays design leveraging knowledge graphs, public literature, and multi-omics data among others.
A step forward in the biodigital industry
The first step towards fully digital cell design, engineering and production