With drug discovery projects costing up to six billion dollars to bring a single drug to market, it's critical that researchers adopt every possible measure to eliminate uncertainties in the development process. Today, one of the most prevailing issues in drug discovery are preclinical animal studies' chronic inconsistency in predicting clinical success. Drugs tested in preclinical models, hailed as the gold standard for translational success, often fail to show efficacy in human trials. This is because the data necessary to improve these models are nearly impossible to aggregate and analyze manually. Our company, ModernVivo, aims to address the barriers preventing evidence-based condition choices by developing a software suite that uses machine learning to help scientists make data-driven decisions on how to design preclinical animal trials. Our technology allows companies to develop studies faster than the status quo while simultaneously reducing experiment development costs.
What is the problem?
Over 90% of drugs that enter human clinical trials have failed to show efficacy, yet every single one of those drugs showed positive results in a preclinical animal study. These translational failures can be traced directly back to the consistent inability of preclinical animal studies to predict clinical success. However, the design process for preclinical animal studies has remained largely unchanged for decades; scientists read a tiny fraction of published literature and write their experimental protocols in roughly a month. This time-consuming, tedious process is not only a major pain point for in-vivo scientists, but causes the unnecessary loss of animal life, millions of dollars in wasted resources, and ineffective drugs entering the clinic.
What is their solution?
When an in vivo scientist develops an experimental protocol today, they embark on a nebulous journey of reading only a small handful of existing published protocols and relying on personal experience. Using ModernVivo instead, in vivo scientists will be guided through the entire experimental design process from choosing which model to run, down to incredibly specific experimental conditions. Furthermore, automating the time-consuming literature review and analysis steps required in this process allows scientists using our software to develop protocols in a fraction of the time (in as little as one hour) that would otherwise be required.
