Researchers unveil new AI models, InstaNovo and InstaNovo+, set to enhance accuracy and discovery in protein science, offering new hope for personalized medicine, drug discovery and more.
In an exciting development poised to transform the landscape of biotechnology and medicine, researchers have introduced two powerful AI models, InstaNovo and InstaNovo+, designed to revolutionize protein science.
This breakthrough comes from a collaborative effort between experts at Technical University of Denmark (DTU), Delft University in the Netherlands and the British AI company InstaDeep. The new models promise to significantly enhance the accuracy of protein sequencing and have the potential to propel advances in personalized medicine, drug discovery and diagnostics.
The advent of widely accessible AI tools has fueled rapid progress across various scientific fields.
In biotechnology, AI-driven innovations are sparking groundbreaking research in drug discovery, precision medicine, gene editing and food security. One crucial sub-field benefiting from these advancements is proteomics — the large-scale study of proteins, the essential molecules that drive biological processes.
Proteomics relies on vast databases that allow scientists to identify proteins and microorganisms in samples, aiding in disease diagnosis, treatment monitoring and pathogen identification.
However, conventional methods have limitations.
“First off, no database includes everything, so you need to know which databases are relevant to your particular needs. Then deep searches are very time-consuming and demand a lot of computer power. And, finally, it’s nearly impossible to identify proteins that haven’t been registered yet,” corresponding author Timothy Patrick Jenkins, an associate professor at DTU Bioengineering, said in a news release
In response to these challenges, researchers have been developing algorithms to enhance accuracy and reduce computational costs. Yet, according to Jenkins and his team, their performance has often been “underwhelming.”
The newly proposed AI models, InstaNovo and InstaNovo+, represent a significant leap forward. These models, featured in a paper published in Nature Machine Intelligence, offer extraordinary precision in de novo peptide sequencing — an essential process in proteomics research.
“Seen together, our models exceed state-of-the-art and are significantly more precise than currently available tools. Furthermore, as we show in the paper, our models are not specific to a particular research area. Instead, these tools could propel significant advances in all fields involving proteomics,” added co-first author Kevin Michael Eloff, a research engineer at InstaDeep.
To validate the models’ effectiveness, the researchers conducted tests across several key areas.
In one study, the models analyzed wound fluid from patients with venous leg ulcers — chronic conditions notoriously resistant to treatment. The models identified 10 times more protein sequences than traditional database searches, including the detection of E. coli and multidrug-resistant Pseudomonas aeruginosa.
In another experiment, the models identified new peptides displayed on cell surfaces, critical for immune system recognition of infections and cancer. This discovery holds enormous promise for personalized cancer immunotherapy, which relies on targeting these peptides.
The implications of their study “extend far beyond the medical sciences,” according to Jenkins.
“Looking at it from a purely technical, scientific perspective, it is also true that with these tools, we can improve our understanding of the biological world as a whole, not only in terms of are but also in industry and academia,” Jenkins added. “Within every field using proteomics — be it plant science, veterinary science, industrial biotech, environmental monitoring, or archaeology — we can gain insights into protein landscapes that have been inaccessible until now.”
The InstaNovo models are already available to researchers through the InstaDeep website, symbolizing a new era of discovery in biological sciences. By enhancing the capabilities of proteomics research, these AI models may unlock unprecedented insights and lead to innovative solutions in various fields, improving people’s lives worldwide.
Source: Technical University of Denmark