Biointron的动态

ICYMI: The enormously challenging work of developing antibodies as therapies may be getting easier. That’s because a new tool developed by GRA Eminent Scholar Jeffrey Skolnick and his Georgia Tech colleague Mu Gao can help predict how certain antibodies will interact with viruses, bacteria and other antigens. They tried the tool on 1,000 potential antibodies for the spike protein of the SARS-CoV-2 virus — and it correctly picked 90% of the best antibodies. The tool is called AF2Complex, and it combines AI and big data to generate deep learning. Check it out: https://bit.ly/4huP3T4

IgGM: A Generative Model for Functional Antibody and Nanobody Design ABSTRACT Immunoglobulins are crucial proteins produced by the immune system to identify and bind to foreign substances, playing an essential role in shielding organisms from infections and diseases. Designing specific antibodies opens new pathways for disease treatment. With the rise of deep learning, AI-driven drug design has become possible, leading to several methods for antibody design. However, many of these approaches require additional conditions that differ from real-world scenarios, making it challenging to incorporate them into existing antibody design processes. Here, we introduce IgGM, generative model that combines a diffusion model and the consistency model for generating antibodies with functional specificity. IgGM produces antibody sequences and structures simultaneously for a given antigen, consisting of three core components: a pre-trained language model for extracting sequence features, a feature learning module for identifying pertinent features, and a prediction module that outputs designed antibody sequences and the predicted complete antibody-antigen complex structure. IgGM has shown effectiveness in both predicting structures and designing novel antibodies and nanobodies, making it relevant in various practical scenarios of antibody and nanobody design. PAPER: https://lnkd.in/ddBFc92A CODE: https://lnkd.in/d9sbNbJA

New AI tool identifies better antibody therapies! From sending cancer into remission to alleviating COVID-19 symptoms, immunotherapy can provide revolutionary disease treatments. Immunotherapies use antibodies—proteins that bind to cell markers called antigens—to target and eliminate the antigen. But despite how effective immunotherapy can be, it isn't widely used because finding the right antibodies to develop treatments is challenging, time-consuming work. Georgia Tech researchers are making this process a little easier, though. Their new tool, AF2Complex, used deep learning to predict which antibodies could bind to COVID-19's infamous spike protein. The researchers created input data for the deep-learning model using sequences of known antigen binders. This method correctly predicted 90% of the best antibodies in one test with 1,000 antibodies and was published in Proceedings of the National Academy of Sciences. Treating COVID-19 is just the start of its potential. https://lnkd.in/gzThJNFp

?? AI-Powered Antibody Revolution: Unlocking New Frontiers in Disease Treatment! ?? Exciting advancements are being made in the field of antibody research! Researchers at MIT have developed a groundbreaking computational technique that leverages large language models (LLMs) to predict antibody structures with unprecedented accuracy. This innovation could pave the way for discovering new antibody drugs that target a wide range of infectious diseases, including SARS-CoV-2. Traditional methods have struggled with the high variability of antibodies, but MIT’s approach focuses on modelling the hypervariable regions of these proteins. As Bonnie Berger, head of the Computation and Biology group at MIT, notes, "Our method allows us to scale... we can actually find a few needles in the haystack." This could significantly reduce the costs associated with clinical trials by helping drug companies avoid pursuing ineffective candidates. By analyzing antibody repertoires from individuals, researchers can gain insights into why certain people respond differently to diseases like HIV and COVID-19. This new model, named AbMap, enables the identification of antibody structures that neutralize viral proteins more effectively than traditional methods. This breakthrough not only enhances our understanding of immune responses but also holds the potential to transform the way we develop treatments for infectious diseases. ?? Let's embrace the future of healthcare powered by AI! #AntibodyDiscovery #AI #ArtificialIntelligence #Biotechnology #InfectiousDiseases #HealthcareInnovation #Research #Science #AIinHealthcare #DrugDevelopment #TheNewClinic https://lnkd.in/gJUSC_uK

** AI in Cancer Treatment AI's impact on healthcare promises faster, more accurate, and scalable solutions to complex medical challenges. These technologies are making drug selection and development faster, more efficient, and cost-effective. I lost my aunt to cancer 12 years ago because of delays in test results and prolonged testing. Many of us have experienced the anguish of watching a loved one suffer from this relentless disease. The advent of AI technology in healthcare could have dramatically altered her prognosis. I am sharing a story of 82-year-old Paul. Diagnosed with aggressive blood cancer that six courses of chemotherapy couldn't eliminate, Paul appeared out of options. Researchers took a tissue sample, divided it into over a hundred pieces, and exposed them to various drug cocktails. Using robotic automation and computer vision, they monitored the reactions. This method tested dozens of treatments simultaneously, avoiding long chemotherapy courses. The AI-driven approach enabled a fast, exhaustive search for the right drug. Traditional methods couldn’t scale this quickly. Two years later, Paul was cancer-free and in complete remission. Nowadays several AI-designed drugs are in clinical trials, undergoing rigorous testing. AI leverages natural-language processing to mine scientific data, identifying promising drug targets and designing novel molecules efficiently. #CancerTreatment #AI #Robotics #HealthcareInnovation #ClinicalTrials #DrugDevelopment #NLP Story reference: (https://lnkd.in/gEfQHRZr)

Over the weekend, I unearthed this engaging paper just published in the ????????????? that explores the transformative potential of artificial intelligence (AI) and machine learning (ML) in noncoding RNA (ncRNA) research. ???????????????? ????????????: ??Role of ncRNAs: The manuscript emphasizes that ncRNAs, which constitute the majority of the human transcriptome, are crucial in regulating genome organization and gene expression at multiple levels, including epigenetic, transcriptional, and post-transcriptional. ??Biomarker Potential: ncRNAs have significant potential as next-generation biomarkers due to their dynamic expression profiles, which can reflect a patient's molecular state and provide insights into disease mechanisms. ??Challenges in Translation: Despite advancements in ncRNA research, translating these findings into clinical practice has been limited by technical and data analysis challenges. Traditional methods often overlook complex interactions between ncRNAs and clinical outcomes. ??Machine Learning Applications: The manuscript highlights the promising role of ML techniques in addressing the biological complexity of ncRNAs. These methods can effectively analyze large, high-dimensional datasets, identifying patterns and relationships that traditional statistical methods might overlook, thereby advancing our understanding of ncRNAs. ??Examples of ML in ncRNA Research: The manuscript provides several examples of ML being successfully applied to identify ncRNA biomarkers, develop diagnostic classifiers, and understand disease mechanisms. For instance, ML models have been used to predict pulmonary arterial hypertension and colorectal cancer prognosis. ??Ethical Considerations: The use of AI/ML in healthcare raises ethical concerns, including data privacy, algorithmic bias, and the need for transparency and trustworthiness in AI models. ??Future Directions: The manuscript underscores the need for collaborative efforts between academia and industry to advance the development of clinically applicable molecular tests. It also suggests that integrating ncRNA data with electronic health records and other omic data (multi-omic strategies) could significantly enhance the clinical utility of ncRNA-based biomarkers. Hope you enjoy it as much as I did! #biotechnology #AI #clinicalresearch #biomarkers https://lnkd.in/dc_EauUP

Accurately predicting antibody structures is essential for developing monoclonal antibodies, pivotal in immune responses and therapeutic applications. Antibodies have two heavy and two light chains, with the variable regions featuring six CDR loops crucial for binding to antigens. The CDRH3 loop presents the greatest challenge due to its diversity. Traditional experimental methods for determining antibody structures are often slow and costly. Consequently, computational techniques such as IgFold, DeepAb, ABlooper, ABodyBuilder, and newer models like xTrimoPGLMAb are emerging as effective tools for precise antibody structure prediction.

Deep learning-based design and experimental validation of a medicine-like human antibody library Abstract Antibody generation requires the use of one or more time-consuming methods, namely animal immunization, and in vitro display technologies. However, the recent availability of large amounts of antibody sequence and structural data in the public domain along with the advent of generative deep learning algorithms raises the possibility of computationally generating novel antibody sequences with desirable developability attributes. Here, we describe a deep learning model for computationally generating libraries of highly human antibody variable regions whose intrinsic physicochemical properties resemble those of the variable regions of the marketed antibody-based biotherapeutics (medicine-likeness). We generated 100000 variable region sequences of antigen-agnostic human antibodies belonging to the IGHV3-IGKV1 germline pair using a training dataset of 31416 human antibodies that satisfied our computational developability criteria. The in-silico generated antibodies recapitulate intrinsic sequence, structural, and physicochemical properties of the training antibodies, and compare favorably with the experimentally measured biophysical attributes of 100 variable regions of marketed and clinical stage antibody-based biotherapeutics. A sample of 51 highly diverse in-silico generated antibodies with >90th percentile medicine-likeness and?>?90% humanness was evaluated by two independent experimental laboratories. Our data show the in-silico generated sequences exhibit high expression, monomer content, and thermal stability along with low hydrophobicity, self-association, and non-specific binding when produced as full-length monoclonal antibodies. The ability to computationally generate developable human antibody libraries is a first step towards enabling in-silico discovery of antibody-based biotherapeutics. These findings are expected to accelerate in-silico discovery of antibody-based biotherapeutics and expand the druggable antigen space to include targets refractory to conventional antibody discovery methods requiring in vitro antigen production. https://lnkd.in/eQjEsSGg

Now for another edition of AI != LLMs (and is actually useful!) There certainly are many posts circulating around at present downplaying the potential and realized possibilities for ML/AI models, where AI has become a term used synonymously with LLM tech such as ChatGPT. That is only a single facet of a much richer space. Here is brand new article discussing use of mRNA based treatment for cancer, which is only possible due to advances in ML/AI algorithms, computational resources, those fancy GPUs and all the rest. AI models are used to personalize antigens on a person by person basis and manufacturing these one off treatments is coordinated and scheduled using optimizations from AI models as well. Amazing work and providing hope for some people suffering from these illnesses. A small group now, but perhaps many more to come as these treatments are showing significant results. In essence they train your immune system to reject and destroy cancer cells without the use of toxic treatments like chemotherapy. https://lnkd.in/ghvPhBKa

In this study, researchers from the University of Toronto present the AI-Guided Ionizable Lipid Engineering (AGILE) platform, a deep learning powered approach to accelerate #LNP development for #mRNA delivery. Ionizable lipid nanoparticles (LNPs) are seeing widespread use in mRNA delivery, notably in SARS-CoV-2 mRNA vaccines. However, the expansion of mRNA therapies beyond COVID-19 is impeded by the absence of LNPs tailored for diverse cell types. #AGILE streamlines ionizable lipid development with efficient library design, in silico lipid screening via deep neural networks, and adaptability to diverse cell lines. Using AGILE, we rapidly design, synthesize, and evaluate ionizable lipids for mRNA delivery, selecting from a vast library. Intriguingly, AGILE reveals cell-specific preferences for ionizable lipids, indicating tailoring for optimal delivery to varying cell types. These highlight AGILE’s potential in expediting the development of customized LNPs, addressing the complex needs of mRNA delivery in clinical practice, thereby broadening the scope and efficacy of mRNA therapies. https://lnkd.in/e2RaE28F #lipids #polymers #lipid #LNPs #lipidnanoparticles #nanoparticles #drugdelivery #DDS #ai #ml #artificialintelligence #machinelearning #biotech #biotechnology #ionizablelipids #cationiclipids