The development of therapeutic antibodies is a complex and resource-intensive endeavor. Traditional methods often involve lengthy processes and high failure rates. But what if we could leverage the power of Artificial Intelligence (AI) to streamline ...
.png)
The development of therapeutic antibodies is a complex and resource-intensive endeavor. Traditional methods often involve lengthy processes and high failure rates. But what if we could leverage the power of Artificial Intelligence (AI) to streamline this process, making it faster, more efficient, and ultimately more successful?
Our recent study published in Frontiers in Immunology explores just that. MAbSilico has developed a pipeline that integrates early-stage phage display screening with AI-based characterization for the design and discovery of novel antibodies against the immune checkpoints TIM3 and TIGIT. Here is a summary of the study’s key findings.
Immune checkpoints like TIM3 and TIGIT are crucial targets in cancer immunotherapy. By blocking these checkpoints, we can unleash the power of the immune system to fight cancer cells.
The study introduces an innovative approach, which is an AI-powered pipeline designed to revolutionize antibody discovery and obtain qualified candidates in less than two months.
Figure 1: Overview of the AI-driven antibody discovery process
The process begins with phage display screening, a robust technique used to identify antibody candidates that exhibit high affinity for the target antigens, TIM3 and TIGIT.
Once potential binders are identified, AI-based tools assess key developability parameters early in the discovery process. This crucial step involves predicting factors that enable researchers to make informed decisions about candidates:
By evaluating these characteristics, AI helps researchers refine the selection of the most promising candidates, reducing failure rates and accelerating development.
AI algorithms are further employed for epitope mapping, enabling the modeling of antibody-target interactions. This step allows researchers to:
This epitope-based refinement enhances both specificity and effectiveness of the selected antibodies.
Recognizing the importance of intellectual property (IP) in drug development, the pipeline also integrates AI-driven IP evaluation. It compares the sequences of the discovered antibodies against a database of existing patented antibodies to determine:
This early-stage assessment secures stronger IP positions and helps avoid potential legal conflicts in future development stages.
The study's integrated approach yielded significant findings that underscore the power of combining phage display with AI-enhanced profiling in antibody discovery.
Phage display screening successfully identified hundreds of binders, which were then refined and selected using AI-based solutions. This process led to the identification of one anti-TIM3 antibody (6E9) and five anti-TIGIT antibodies (T2, T4, T7, and T10).
These selected antibodies initially displayed:
The incorporation of AI-driven developability assessments provided critical insights for candidate selection.
T2 and T10 were flagged for intense hydrophobic regions, a characteristic known to promote aggregation and potentially impede their progression as viable therapeutic agents.
Early identification of developability issues allowed for a more informed selection process.
The structural modeling further refined this process by providing valuable structural information for T4 and 6E9, ultimately leading to their prioritization as the top candidates against TIGIT and TIM3, respectively.
Furthermore, the study validated the effectiveness of the affinity method in accurately ranking binders, a result that supports:
Figure 2: Aggregation parameters analysis.
This study significantly advances the field by presenting a pioneering approach that integrates AI-based tools for developability and epitope prediction to streamline and accelerate antibody discovery.
The identification of novel TIM3 and TIGIT antibodies holds significant promise for developing the next generation of cancer immunotherapies. Their comprehensive characterization—including binding properties, developability profiles, and predicted epitopes—provides essential insights to guide further development and optimization.
Looking ahead, future research endeavors will be geared toward thorough investigation:
By leveraging AI-powered discovery and optimization, our approach accelerates the path to effective cancer treatments, highlighting the role of AI in revolutionizing biopharmaceutical innovation.
Reference(s):
Front. Immunol. , 11 March 2025, Sec. Cancer Immunity and Immunotherapy, Volume 16 - 2025 | Frontiers | AI-enhanced profiling of phage-display-identified anti-TIM3 and anti-TIGIT novel antibodies