Screen more biological possibilities before you run more experiments.
BioSymetrics’ virtual phenotypic screening platform uses human and model-system phenotypes to predict novel gene targets in silico, rank the hypotheses worth testing, and move them into high-throughput in vivo validation.
Start with the biology you want to change.
Disease biology is complex. A phenotype can show that something is happening without showing which gene or pathway to test next. Finding those connections often means working through large experimental matrices before the most useful hypotheses emerge.
Our virtual screening workflow helps focus that search. It starts with the biological effect of interest and uses machine learning to identify the gene perturbations most likely to explain or change it.
Bring clinical, genomic, and experimental signals into one view.
We combine clinical, genomic, and experimental phenotype data to build a machine-learning representation of the biology being screened. Signals from human disease and model systems can then be compared and translated into candidate gene-phenotype relationships.
These candidate gene-phenotype relationships give target discovery a phenotype-led starting point grounded in both human biology and experimental evidence.
Rank gene perturbations before committing to the bench.
Our models predict the effects of disrupting thousands of genes in silico. The result is a ranked set of target and therapeutic hypotheses that helps determine which experiments to run first.
Validation remains part of the workflow. The virtual screen narrows the search space so experimental work can focus on the hypotheses with the clearest biological rationale.
Take the strongest hypotheses into living systems.
We test shortlisted predictions in CRISPR-modified zebrafish using scalable 96-well experiments. Zebrafish develop quickly, are straightforward to manipulate genetically, and provide a practical model for testing biological effects in vivo.
High-content phenotypic readouts can be generated within days of gene manipulation, allowing results from validation to feed into the next round of discovery.
Turn in vivo responses into rich phenotypic data.
Automated imaging and computer vision quantify changes across morphology, blood flow, heart function, and behaviour. These readouts capture anatomy, physiology, and activity in the same experimental workflow.
The result is dense, quantitative evidence that supports target validation and hit identification at high throughput.
Move from ranked targets to the next experiment.
Screening outputs support target prioritization, target validation, hit identification, biomarker discovery, and follow-up mechanism work. They can help partners decide which biological hypotheses to advance and which experiments can provide the most useful next evidence.
BioSymetrics applies this approach across cardiac, neurodegenerative, and rare diseases.
Interested in seeing what virtual phenotypic screening could reveal in your discovery program?
We work with pharmaceutical, biotech, hospital, and translational research partners to advance existing discovery programs or begin new ones in cardiac, neurodegenerative, and rare diseases.
Partner with us