Revolutionizing Antibody Design: Harnessing the Power of Generative AI

Introduction

The field of generative artificial intelligence has seen remarkable advancements in recent years, enabling machines to create images and generate text. However, its potential goes far beyond art and literature. Absci Corporation, a pioneering company in antibody therapeutics, is leveraging generative AI to revolutionize drug discovery. By harnessing the power of machine learning, Absci can create novel antibodies, a breakthrough that can potentially transform the field of medicine. Absci has achieved numerous milestones and garnered significant recognition. The company has secured over $425 million in funding as a publicly traded company on Nasdaq. With a team of more than 200 employees, Absci has solidified its position as a powerhouse in the biotech industry. These acquisitions have strengthened Absci’s capabilities and expanded its offerings. This article delves into Absci’s innovative approach, exploring how generative AI shapes the future of antibody design.

Absci’s Journey and Expanding Capabilities

Founded in 2011, Absci Corporation has experienced a rapid evolution, transforming from a small lab into a massive facility spanning 77,400 square feet in Vancouver, Washington. The company has also expanded its operations with additional New York City and Switzerland sites. Collaborating with top antibody therapeutics manufacturers worldwide, Absci has developed cutting-edge screening capabilities. Those capabilities allow them to test and validate nearly 3 million unique AI-generated antibody designs weekly. In addition, they have recently introduced an AI-powered computational platform called the Integrated Drug Creation™ platform. This platform enables the analysis of over a billion molecules computationally per week. More than 100,000 of these molecules are being validated in the laboratory.

The Role of AI in Biology and Medicine

Just as AI has revolutionized various domains, including information and technology, it is now making significant breakthroughs in biology and medicine. Absci recognized this opportunity early and hired Joshua Meier, formerly of Meta, to lead their de novo protein design efforts using large language models. As the Chief AI Officer at Absci, Meier’s role highlights AI’s transformative potential in antibody therapeutics. Meier states that they are achieving feats that were once deemed impossible just a few years ago. He also highlights how a programmer can now save lives with code written 18-24 months prior.

The Challenge of Protein Design and Antibodies

Proteins, often called biological machines, play essential roles in our bodies by executing various bodily functions. The DNA sequence of a gene encodes the information necessary for a protein to perform its specific tasks. Designing proteins to fulfill particular functions requires understanding the intricate relationship between protein sequence and function. Among the proteins, antibodies pose one of the most challenging design problems. Antibodies play a crucial role in our immune system, and they can be engineered to bind to specific tissues, fight cancer cells, or neutralize infectious viruses. However, achieving precise binding is notoriously difficult, and even the most promising designs often fail in clinical trials.

Absci’s Approach to Antibody Design

Absci has been at the forefront of tackling the complexities of antibody design for a considerable period. They have assisted their clients in developing improved therapies that exhibit high specificity, stability within the body, and a reduced likelihood of triggering an immune response. Based on this expertise, Absci uses generative AI to develop its antibody pipeline. AI-aided protein design requires vast amounts of training data to teach the models. Unfortunately, the field of antibody development has struggled to provide the high-quality biological data necessary to generate reliable AI models. Recognizing this challenge, Absci has spent the past decade collecting antibody data, enabling them to develop their zero-shot generative AI method.

The Significance of Zero-Shot Generative AI

“Zero-shot” refers to the ability to design antibodies to bind to specific targets without relying on training data from known antibodies with similar binding characteristics. Absci’s zero-shot approach ensures more sequence diversity in de novo designs and eliminates bias. By eschewing training data for specific targets, the generative AI models can explore uncharted territories and provide innovative solutions to antibody design. This method could overcome the limitations of existing antibody therapies and revolutionize the treatment landscape.

Absci Corporation’s integration of generative AI into antibody design represents a significant milestone in drug discovery. With its cutting-edge computational and wet lab capabilities, Absci is at the forefront of leveraging AI to create novel antibodies. By combining machine learning algorithms with their extensive collection of antibody data, they are pushing the boundaries of what is possible in antibody therapeutics. As generative AI continues to evolve, the potential for accelerating the discovery and development of life-saving treatments becomes increasingly promising. The breakthroughs achieved by Absci and similar companies in this space highlight the transformative power of AI in medicine and offer hope for improving the lives of countless individuals in need of innovative therapies.