Summary Bettering Human Health Through Artificial Intelligence with Sean McClain and Joshua Meier of Absci (Youtube) www.youtube.com
11,520 words - YouTube video - View YouTube video
One Line
Absci is leveraging generative AI to transform biologic drug discovery by improving success rates and reducing time and costs.
Slides
Slide Presentation (12 slides)
Key Points
- Absci is utilizing generative AI to revolutionize biologic drug discovery.
- Their approach involves engineering E. coli to produce antibodies, enabling the creation of a billion-member antibody library for screening.
- Generative AI allows for the design of antibodies with desired attributes, transforming the paradigm from drug discovery to drug creation.
- Absci's integration of AI and biology allows them to generate AI-designed data, providing a robust training set for their models.
- Their optimization efforts focus on creating antibodies that are developable and manufacturable while minimizing immunogenicity.
- Absci aims to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market through AI optimization.
- Collaboration and sharing of data among industry players could help address unknown unknowns in drug discovery.
- Absci sees the decreasing cost of data generation as a positive development that will lead to greater efficiency in drug discovery and more strategic decision-making.
Summaries
19 word summary
Absci is using generative AI to revolutionize biologic drug discovery, increasing the success rate and reducing time and cost.
114 word summary
Absci, a public company, is using generative AI to revolutionize biologic drug discovery. Founder and CEO Sean McClain and Chief AI Officer Joshua Meier discuss the role of AI in drug development and its impact on creating faster and more effective medicines at a lower cost. Absci's approach involves engineering E. coli to create a billion-member antibody library for screening, leading to faster and more accurate results. By training AI models using billions of data points, Absci can rapidly screen and validate millions of AI-generated designs within six weeks. Their work has the potential to increase the success rate of drug candidates and reduce the time and cost of bringing new drugs to market.
144 word summary
Absci, a public company, is revolutionizing biologic drug discovery using generative AI. Founder and CEO Sean McClain and Chief AI Officer Joshua Meier discuss the role of AI in drug development and its impact on creating more effective medicines faster and at a lower cost. Absci's journey began 12 years ago, initially focusing on engineering E. coli to produce antibodies. Three years ago, they realized the potential of generative AI in unlocking biologic drug discovery. Absci's approach involves engineering E. coli to create a billion-member antibody library for screening, leading to faster and more accurate results. By training AI models using billions of data points, Absci can rapidly screen and validate millions of AI-generated designs within six weeks. Their work has the potential to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market.
429 word summary
Absci, a public company, is utilizing generative AI to revolutionize biologic drug discovery. The company's founder and CEO, Sean McClain, and Chief AI Officer, Joshua Meier, discuss the role of AI in drug development and the impact it has on creating more effective medicines faster and at a lower cost.
Absci's journey began 12 years ago when McClain founded the company in a basement lab. Initially, their focus was on engineering E. coli to produce antibodies. However, it wasn't until three years ago that McClain realized the potential of generative AI in unlocking biologic drug discovery.
Absci's approach involves engineering E. coli to produce antibodies, enabling the creation of a billion-member antibody library for screening. This method offers control over the generation of antibodies, leading to faster and more accurate results.
The key challenge in antibody discovery is achieving specificity and affinity. Absci's use of generative AI allows for the design of antibodies with desired attributes, transforming the paradigm from drug discovery to drug creation.
Absci's innovative approach involves training AI models using billions of data points generated from their microbial platform. This data-driven approach allows for the rapid screening and validation of millions of AI-generated designs within a six-week timeframe.
Absci's integration of AI and biology allows them to generate AI-designed data, providing a robust training set for their models. By leveraging AI to generate and validate their designs, Absci has achieved remarkable results in a short timeframe.
Absci's optimization efforts go beyond affinity and focus on creating antibodies that are developable and manufacturable while minimizing immunogenicity. The naturalness model developed by Absci correlates inversely with immunogenicity and provides reliable metrics for developability and manufacturability.
Absci, a company specializing in artificial intelligence (AI) for drug discovery, aims to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market.
Absci utilizes AI models to generate potential sequences that could be promising drug candidates. To determine which molecules to prioritize, they incorporate information about naturalness, affinity, and other properties.
Absci believes that personalized medicine is the future of healthcare. This shift towards personalized medicine requires a deep understanding of biology, which AI models can help provide.
In conclusion, Absci is at the forefront of using AI to revolutionize drug discovery. By leveraging AI models, incorporating data-driven approaches, and focusing on creating developable antibodies, Absci has achieved remarkable results in a short timeframe. Their work in AI has the potential to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market.
657 word summary
Absci, a public company, is utilizing generative AI to revolutionize biologic drug discovery. The company's founder and CEO, Sean McClain, and Chief AI Officer, Joshua Meier, discuss the role of AI in drug development and the impact it has on creating more effective medicines faster and at a lower cost.
Absci's journey began 12 years ago when McClain founded the company in a basement lab. Initially, their focus was on engineering E. coli to produce antibodies. This allowed them to create a large library of antibodies for high-throughput screening. However, it wasn't until three years ago that McClain realized the potential of generative AI in unlocking biologic drug discovery.
To understand the importance of Absci's work, it is essential to grasp the process of antibody discovery and development. Absci's approach involves engineering E. coli to produce antibodies, enabling the creation of a billion-member antibody library for screening. This method offers control over the generation of antibodies, leading to faster and more accurate results.
The key challenge in antibody discovery is achieving specificity and affinity. Currently, antibodies are generated using animal models or phage display techniques. However, these methods lack control over the generation of antibodies with specific attributes, resulting in long lead times and low success rates in clinical trials. Absci's use of generative AI allows for the design of antibodies with desired attributes, such as epitope specificity and binding affinity, thus transforming the paradigm from drug discovery to drug creation.
Absci's innovative approach involves training AI models using billions of data points generated from their microbial platform. They utilize a breakthrough assay called ACE that can interrogate every E. coli in their library and assess binding affinity. This data-driven approach allows for the rapid screening and validation of millions of AI-generated designs within a six-week timeframe.
The importance of data in AI-enabled drug discovery cannot be overstated. Absci's integration of AI and biology allows them to generate AI-designed data, providing a robust training set for their models. This approach enables the rapid scaling of experiments, leading to insights and intuition about what works and what doesn't. By leveraging AI to generate and validate their designs, Absci has achieved remarkable results in a short timeframe.
Absci's optimization efforts go beyond affinity and focus on creating antibodies that are developable and manufacturable while minimizing immunogenicity. The naturalness model developed by Absci correlates inversely with immunogenicity and provides reliable metrics for developability and manufacturability. By training their models on hundreds of millions of naturally occurring antibody sequences, Absci ensures that their designs align with human immune repertoires.
Absci, a company specializing in artificial intelligence (AI) for drug discovery, is focused on improving human health through innovative technology. They use AI models to optimize the selection and design of molecules for drug development. By applying AI to the drug discovery process, Absci aims to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market.
One of the key challenges in drug discovery is prioritizing molecules for further testing. Absci utilizes AI models to generate potential sequences that could be promising drug candidates. To determine which molecules to prioritize, they incorporate information about naturalness, affinity, and other properties.
Absci draws parallels between their work in AI and the field of natural language processing (NLP). Just as language models learn from large amounts of text data, AI models in drug discovery learn from vast amounts of biological data. The analogy between proteins and language is particularly interesting, with proteins having 20 canonical amino acids and English having 26 characters.
Absci believes that personalized medicine is the future of healthcare. As the success rate of drug development increases, it becomes economically viable to target smaller patient populations. This shift towards personalized medicine requires a deep understanding of biology, which AI models can help provide.
In conclusion, Absci is at the forefront of using AI to revolutionize drug discovery. By leveraging AI models, incorporating
1284 word summary
Absci, a public company, is utilizing generative AI to revolutionize biologic drug discovery. The company's founder and CEO, Sean McClain, and Chief AI Officer, Joshua Meier, discuss the role of AI in drug development and the impact it has on creating more effective medicines faster and at a lower cost.
Absci's journey began 12 years ago when McClain founded the company in a basement lab. Initially, their focus was on engineering E. coli to produce antibodies. This allowed them to create a large library of antibodies for high-throughput screening. However, it wasn't until three years ago that McClain realized the potential of generative AI in unlocking biologic drug discovery.
To understand the importance of Absci's work, it is essential to grasp the process of antibody discovery and development. Traditionally, antibodies are produced in mammalian or CHO cells, which have limitations in terms of scalability and data generation. Absci's approach involves engineering E. coli to produce antibodies, enabling the creation of a billion-member antibody library for screening. This method offers control over the generation of antibodies, leading to faster and more accurate results.
The key challenge in antibody discovery is achieving specificity and affinity. Currently, antibodies are generated using animal models or phage display techniques. However, these methods lack control over the generation of antibodies with specific attributes, resulting in long lead times and low success rates in clinical trials. Absci's use of generative AI allows for the design of antibodies with desired attributes, such as epitope specificity and binding affinity, thus transforming the paradigm from drug discovery to drug creation.
Absci's innovative approach involves training AI models using billions of data points generated from their microbial platform. They utilize a breakthrough assay called ACE that can interrogate every E. coli in their library and assess binding affinity. This data-driven approach allows for the rapid screening and validation of millions of AI-generated designs within a six-week timeframe. The combination of AI and wet lab validation enables Absci to make significant progress in biologic drug discovery.
The importance of data in AI-enabled drug discovery cannot be overstated. Absci's integration of AI and biology allows them to generate AI-designed data, providing a robust training set for their models. This approach enables the rapid scaling of experiments, leading to insights and intuition about what works and what doesn't. By leveraging AI to generate and validate their designs, Absci has achieved remarkable results in a short timeframe.
When considering AI and drug discovery, it is crucial to identify red flags and green flags that indicate differentiated and valuable approaches. Wet lab validation is a key factor in distinguishing legitimate advancements from hype. Absci's focus on epitope specificity, accuracy in predicting binding affinity, and model accuracy are critical metrics for success. Additionally, the ability to unlock new biology, such as targeting GPCRs, further demonstrates the potential of generative AI in biologic drug discovery.
Absci's optimization efforts go beyond affinity and focus on creating antibodies that are developable and manufacturable while minimizing immunogenicity. The naturalness model developed by Absci correlates inversely with immunogenicity and provides reliable metrics for developability and manufacturability. By training their models on hundreds of millions of naturally occurring antibody sequences, Absci ensures that their designs align with human immune repertoires.
In conclusion, Absci's integration of generative AI and biology has revolutionized biologic drug discovery. Their use of AI models trained on billions of data points allows for rapid screening and validation of AI-generated designs. The focus on epitope specificity, binding affinity, and naturalness ensures the development of effective and safe antibodies. With the ability to unlock new biology and target challenging areas like GPCRs, Absci is at the forefront of innovation in the field.
Absci, a company specializing in artificial intelligence (AI) for drug discovery, is focused on improving human health through innovative technology. They use AI models to optimize the selection and design of molecules for drug development. By applying AI to the drug discovery process, Absci aims to increase the success rate of drug candidates and reduce the time and cost associated with bringing new drugs to market.
One of the key challenges in drug discovery is prioritizing molecules for further testing. Absci utilizes AI models to generate hundreds or even thousands of potential sequences that could be promising drug candidates. To determine which molecules to prioritize, they incorporate information about naturalness, affinity, and other properties. This allows them to select the most interesting molecules for preclinical testing.
Another approach Absci takes is to optimize lead molecules for various properties from the start. By combining different properties and using AI models, they can co-optimize sequences to achieve desired affinity and natural profiles. This flexibility in combining properties enables Absci to tailor their approach based on the specific problem they are trying to solve.
In drug discovery, it is crucial to eliminate as many unknowns and potential pitfalls as possible in order to increase the chances of a drug successfully reaching commercialization. While many known unknowns can be addressed through careful optimization and testing, there are still unknown unknowns that can pose challenges. For example, post-translational modifications that are not directly measured by current techniques could impact the binding properties of antibodies. Absci acknowledges the importance of addressing these unknown unknowns and suggests that collaboration and sharing of data among industry players could help overcome these challenges.
Absci draws parallels between their work in AI and the field of natural language processing (NLP). Just as language models learn from large amounts of text data, AI models in drug discovery learn from vast amounts of biological data. The analogy between proteins and language is particularly interesting, with proteins having 20 canonical amino acids and English having 26 characters. Absci emphasizes the importance of incorporating meaningful information about target protein structures into their AI models to enhance their performance.
While Absci's initial focus is on antibodies, they recognize that their AI models can be applied to other biologics and different modalities. They highlight the need to adapt the models to the unique characteristics of each modality. For example, antibodies primarily rely on the binding conferred by complementarity-determining regions (CDRs), while general proteins require models that can represent a broader range of information. As the scope expands to include small molecules, protein dynamics, and whole cells, Absci emphasizes the importance of building the right biases into the models to capture the relevant information.
The cost of generating data points for AI models is decreasing over time, thanks to advancements in DNA synthesis and sequencing technologies. This trend enables more scalable and cost-effective data generation, making it possible to produce larger amounts of data for training models. Absci sees this as a positive development that will lead to greater efficiency in drug discovery and more strategic decision-making. They envision a future where scientists can trust AI models to guide their research and focus on the real-world applications of their work.
Absci believes that personalized medicine is the future of healthcare. As the success rate of drug development increases, it becomes economically viable to target smaller patient populations. This shift towards personalized medicine requires a deep understanding of biology, which AI models can help provide. By training models on vast amounts of data and incorporating synthetic biology technologies, Absci aims to design drugs that not only hit the desired targets but also achieve the desired biological effects.
In conclusion, Absci is at the forefront of using AI to revolutionize drug discovery. By leveraging AI models, incorporating meaningful information, and embracing synthetic biology, they aim to improve the success rate of drug candidates and accelerate the development of personalized medicine. The decreasing cost of data generation and the potential for commoditization in the field create exciting opportunities for advancements in drug discovery.
Raw indexed text (63,706 chars / 11,520 words)
Source: https://www.youtube.com/watch?v=q5GEs2MYKek
Page title: Bettering Human Health Through Artificial Intelligence with Sean McClain and Joshua Meier of Absci - YouTube
Meta description: Drug development is an extremely expensive endeavor, in large part because of the 96% failure rate of clinical drug trials. However, the new frontier of medi...