In the “Golden Age” of antibiotic discovery during the 1950s and 60s, pharmaceutical companies, research institutes and scientists joined the forces in a global hunting for life -saving compounds. He scattered soil samples from Amazon to the Himalayas, exchanged microbial strains in institutions, and shared shared discoveries – such as streptomycin, first effective treatment for tuberculosis – within weeks, a pre -unprecedented speed. This was a time when science progressed rapidly through remarkable openness and cooperation.
In decades, antibiotic resistance has increased in a fully developed crisis-not for scientific information or lack of effort, but due to deep structural challenges. Nevertheless, the associate spirit of that era provides a powerful model of how we can respond.
According to the World Health Organization, there is anti -antimicrobial resistance One of the top 10 global public health hazards facing humanityNevertheless, only a handful of antibiotics have been approved in recent years. With limited financial reverse, small, mission-run teams are now carrying torches for antibiotic search-a fraction of masculinity and increasing demand for high quality data. Today, only three major pharmaceutical companies continue to maintain antibiotic search programs, as the vast majority are out of the field.
Most earn new antibiotics Less than $ 50 million per yearDespite treating millions. Meanwhile, other areas of the drug can generate billions of annually annually, often serving small patient populations. This inequality highlights how market incentives do not always align with essential public health requirements.
Already, antibiotic is responsible for resistance More than 1.27 million deaths annually Overall, contributing to about 5 million deaths – and without intervention, it becomes one of the major causes of death worldwide by 2050.
AI makes a real promise to reversed this trend. By quickly generating novel compounds in a rapid response loop with experimental verification, AI can cut both time and cost – reducing failed experiments and enabling less, but better, “shots on the target.” But even the most advanced algorithm cannot give results without some founders: data.
Despite the AI capacity, most small biotech face a malignant obstruction: accessible, high quality data deficiency. Training effective AI model requires tens of thousands of data points, each costs around $ 50. It is only $ 5 million to just start. For startups working in low-marginal areas such as antibiotics, it is often out of access.
In Fair Bio, we have experienced these hurdles.
In our early days, we rely on the AI model, a future AI model to identify antibiotic properties in existing libraries of millions of compounds. It was during this period that our educational partners in the Collins Lab of MIT discovered the first novel antibiotic class using AI.
But still, our ability to generate promising lead was beating the available data to support meaningful discovery. Many of the datasets we trusted were already heavy mining by other researchers, which limited the innovation of the conclusions and the capacity for new intellectual property.
Today, we are charting our own courses using generic AIs to fully propose new chemical structures-novels such as the world’s deadliest superbugs, drugs such as compounds. This approach has already given rise to two additional novel antibiotic classes, which accelerates our speed, accuracy and scale. Nevertheless, we still have a lack of data to maintain our model’s ability.
Therefore, with government and philanthropic support, we are generating our own datasets, collecting public data, and building lean workflows to spread every dollar. We have created a viable economic path-and we are committed to opening the data that generates others to help others. But we can move even faster if we can galvanize comprehensive data-sharing efforts throughout the region.
Data sharing does not mean control. In fact, AI and Cloud technologies have made it easier to cooperate responsibly than ever. Federated learning allows model training without leaving its source. Time-limited agreements and anonymous training protocols help ensure safety without highlighting ownership data.
We have already seen how safe, coordinated data-sharing can accelerate scientific progress. For example, the protein data bank has enabled the successes of decades by creating structural data on independently available proteins to researchers worldwide.
Imagine a world where biotech can safely train better models using each other’s data – without it, except for individual platforms and how it is used, with complete transparency about it. Where large pharmaceutical companies lend to small biotech inactive or non-sensitive data or license, especially in areas with limited financial inverted or frequent scientific intervals, promotes the discovery of new drug, which occurs in medical fields of unmatched needs.
Unlike traditional partnerships that are often in the drug growth process and behind closed doors, this model is attracted to open science principles-adequate-phase data-sharing, decentralized collaboration and promote search to encourage shared infrastructure. In turn, pharma companies can license the most promising compounds and can take them through pre -of pre -clinical and clinical development – or support Biotech in advancing them. Startups get data and capital to move forward. Pharma re-connects with de-stiff assets in once refined areas. And most importantly, patients get immediate requirement, rapid access to life -saving drugs.
We are fighting the 21st century battle with the 20th -century equipment. As we rapidly increase machine and human intelligence to powerful effects, we are in a transformational moment where open science can once again unlock a marine change in biotechnology.
We should revive the golden age of drug discovery: collaborative, practical, immediate. The bets are too much to keep quiet.
Data is out of there. Let us use it to save life.
Akhila Kosarju, MD, Fare Bio is the president and CEO , The novel takes advantage of a biotech social venture liberal AI to discover the next generation of antibiotics.
