Why does Alltrna need #machinelearning to unlock #tRNA biology? tRNAs are programmable molecules with a diverse biology of sequences and modifications that are key to their structure, function, and stability. There’s the potential for approximately 10^34 tRNA sequences and more than 120 natural and synthetic modifications for each nucleotide – yielding the opportunity to generate more engineered tRNA oligonucleotides than atoms in the universe What this diversity means is that there is a huge amount of flexibility in tRNA sequences without perturbing their function. Our ML-enabled platform empowers Alltrna to explore the possible sequence space to design therapeutic molecules optimized not only for biological activity, but also for tolerability, stability, manufacturability, and a host of other characteristics required in drug design. Our CEO Michelle Werner will explore this topic and more at the #RNALeaders conference in #SanDiego next month. Hear a preview of the talk here: https://lnkd.in/gPqgNu3R See the RNA Leaders conference agenda here – note new day/time for the talk: https://lnkd.in/g93QgSzv #RNA #tRNAmedicines #tRNAbiology #rarediseases #geneticmedicines #programmablemedicines #AI #machinelearning #StopCodonDisease
Alltrna’s Post
More Relevant Posts
-
Great webinar series ahead focusing on targeting RNA. Be on the edge of science and register for more insights. #Webinar #RNA #Modulation #SmallMolecule #DrugDiscovery
[UPCOMING WEBINAR 💻] "How to target RNA modulation with Small Molecule (Part 1) 🧬 Discover our speakers : 👩🔬 Dr. Ella Morishita, senior investigator at Veritas In Silico Inc., will present: « Probing RNA−Small Molecule Interactions Using Biophysical and Computational Approaches” 👩🔬 Dr. Maria Duca, Research Scientist at UNIVERSITE DE NICE, will present: « Targeting of non-coding RNAs using synthetic small molecules » Register now for Part 1️⃣👉https://lnkd.in/eU_n5jwf #Webinar #RNA #Modulation #SmallMolecule #DrugDiscovery
-
Professor & Research Director @ Novo Nordisk Foundation Center for Protein Research | PhD in Chemistry
Sharing our experience with Alphafold3 after determining the cryo-EM structure of Retron-Eco1 See Molecular Cell 👇 https://lnkd.in/drdeCKNP We tested Alphafold3 in this protein-nucleic acids complex. Retron-Eco1 is composed by a reverse transcriptase, an effector N-glycosidase, msDNA and RNA. The effector degrades NAD+ and is encapsulated by the msDNA. The attached video shows the top Alphafold3 model compared to the structure of the minimal segment of the experimental filamentous assembly of Retron-Eco1. Perhaps "professional alphafolders" could get better results. Anyway, as it can be observed in the video the prediction is far from reflecting the structure of Retron-Eco1. The protein folds are OK but the assembly in the nucleic acids scaffolding is far from the experimental structure. It is important to mention that Retron-Eco1 the assembly is designed by Nature to stabilize ADPr from NAD+ hydrolysis covalently linked to the catalytic E106 (see fig 4B in the paper). We did not know this feature, neither did Alphafold3. This type of important information leading to new discoveries are difficult to include in the data used to train the program. Don´t misunderstand me, I think Alphafold is a precious invaluable “hypothesis” generator tool that has accelerated our work tremendously, however, it must be handled with care, as important aspects are not grasped by the AI. This could bias our view, compromising new findings. #cryoem #structuralbiology #alphafold #antimicrobials #phagetherapy
-
Exciting breakthrough in #cellbiology! 🔬 Scientists propose AI-powered Virtual Cells to revolutionize our understanding of biology and disease. These models learn from vast biological data to create robust representations of cells under various conditions. 🎯 This groundbreaking approach could transform drug discovery, predict cellular responses, and accelerate hypothesis testing. Quick Read: https://lnkd.in/gmUrbtAQ #Bioinformatics #VirtualCell #ArtificialIntelligence #BiomedicalResearch #AIinBiology #AIinHealthcare #ScienceNews
-
-
Be careful with your theoretical models... 🤓 While AlphaFold represents a significant advancement in contemporary scientific research, it is not without its limitations.😁🫡 The accuracy and resolution of its theoretical models can be compromised by insufficient and suboptimal sampling of certain types of molecular structures. 🤔😯 This limitation underscores the necessity for continued empirical data collection and validation to enhance the reliability and applicability of AlphaFold-generated predictions in diverse molecular contexts.🧑💻🤯
Professor & Research Director @ Novo Nordisk Foundation Center for Protein Research | PhD in Chemistry
Sharing our experience with Alphafold3 after determining the cryo-EM structure of Retron-Eco1 See Molecular Cell 👇 https://lnkd.in/drdeCKNP We tested Alphafold3 in this protein-nucleic acids complex. Retron-Eco1 is composed by a reverse transcriptase, an effector N-glycosidase, msDNA and RNA. The effector degrades NAD+ and is encapsulated by the msDNA. The attached video shows the top Alphafold3 model compared to the structure of the minimal segment of the experimental filamentous assembly of Retron-Eco1. Perhaps "professional alphafolders" could get better results. Anyway, as it can be observed in the video the prediction is far from reflecting the structure of Retron-Eco1. The protein folds are OK but the assembly in the nucleic acids scaffolding is far from the experimental structure. It is important to mention that Retron-Eco1 the assembly is designed by Nature to stabilize ADPr from NAD+ hydrolysis covalently linked to the catalytic E106 (see fig 4B in the paper). We did not know this feature, neither did Alphafold3. This type of important information leading to new discoveries are difficult to include in the data used to train the program. Don´t misunderstand me, I think Alphafold is a precious invaluable “hypothesis” generator tool that has accelerated our work tremendously, however, it must be handled with care, as important aspects are not grasped by the AI. This could bias our view, compromising new findings. #cryoem #structuralbiology #alphafold #antimicrobials #phagetherapy
-
I like to think of AlphaFold as one of the best applications of machine learning to date. It gives us a tremendous tool to understand a problem that was previously intractable without machine learning, but it ultimately suffers from the same problems that all modeling software does: it really works best for interpolation. This post really highlights the risks of relying entirely on AI-generated "data" (more accurately, predictions). This result shows that ML is still not always a substitute for human-driven research and experimental data.
Professor & Research Director @ Novo Nordisk Foundation Center for Protein Research | PhD in Chemistry
Sharing our experience with Alphafold3 after determining the cryo-EM structure of Retron-Eco1 See Molecular Cell 👇 https://lnkd.in/drdeCKNP We tested Alphafold3 in this protein-nucleic acids complex. Retron-Eco1 is composed by a reverse transcriptase, an effector N-glycosidase, msDNA and RNA. The effector degrades NAD+ and is encapsulated by the msDNA. The attached video shows the top Alphafold3 model compared to the structure of the minimal segment of the experimental filamentous assembly of Retron-Eco1. Perhaps "professional alphafolders" could get better results. Anyway, as it can be observed in the video the prediction is far from reflecting the structure of Retron-Eco1. The protein folds are OK but the assembly in the nucleic acids scaffolding is far from the experimental structure. It is important to mention that Retron-Eco1 the assembly is designed by Nature to stabilize ADPr from NAD+ hydrolysis covalently linked to the catalytic E106 (see fig 4B in the paper). We did not know this feature, neither did Alphafold3. This type of important information leading to new discoveries are difficult to include in the data used to train the program. Don´t misunderstand me, I think Alphafold is a precious invaluable “hypothesis” generator tool that has accelerated our work tremendously, however, it must be handled with care, as important aspects are not grasped by the AI. This could bias our view, compromising new findings. #cryoem #structuralbiology #alphafold #antimicrobials #phagetherapy
-
World DNA Day. Myself and our team take pride in pushing the boundaries of Next Generation Sequencing with AI technology. More about Geneyx Analysis here >https://lnkd.in/dRHtBVHz #DNAday #WDD2024 #genomics #genetics #biotechnology
25th April is “World DNA Day”. 🧬🧬 This years theme aims to promote progress in biotechnology with the introduction of advanced intelligence in large quantities. 🔹Why now? WDD - 2024 commemorates the day in 1953 when James Watson, Francis Crick, Maurice Wilkins, Rosalind Franklin and colleagues published papers in the journal Nature Portfolio on the structure of DNA. 🔹To understand how AI is changing Next Generation Sequencing data to identify novel biomedical insights, while also improving diagnostic yields and turnaround times ➡️ take a look at Geneyx Analysis. 💡https://lnkd.in/dGFJYbCB #WDD2024 #dna #rna #NGS #hospitals
-
-
A recently discovered variant of type I CRISPR-Cas systems, HNH-Cascade, lacks a Cas3 catalytic module and instead uses an HNH endonuclease domain to degrade DNA. To explore how HNH-Cascade works, Seiichi Hirano, Feng Z., and colleagues resolved the cryo-EM structure of HNH-Cascade from Selenomonas bacteria together with its target DNA. They observed that this Cascade scaffold adapted to work with the unique HNH insertion by forming a ring-like architecture that enables precise DNA binding and cleavage. The results highlight how these adaptable nuclease domains achieve diverse functions. Read more in Molecular Cell. #BroadInstitute #Science #ScienceNews #Research #ScientificResearch
Structural determinants of DNA cleavage by a CRISPR HNH-Cascade system
cell.com
-
📍 Spatial domain detection using contrastive self-supervised learning for spatial multi-omics technologies Yao et al. introduced Proust, a scalable tool to predict discrete domains using spatial multi-omics data by combining the low-dimensional representation of biological profiles based on graph-based contrastive self-supervised learning. Key points ✔ Integration of multiple data modalities, such as RNA, protein, and H&E images ✔ Prediction of spatial domains within tissue samples ✔ Enhanced accuracy as evidenced across various benchmark datasets and technological platforms ➡ More details: https://lnkd.in/eJSaTHQZ #JohnsHopkinsBloombergSchoolofPublicHealth #spatialomics #spatialbiology #singlecellanalysis #singlecell
-
-
Happy Monday everyone! Today's blog dives deeper into a new method - published in Nature - called Slide-tags. It's changing spatial omics by capturing single-cell data with high-resolution location information. Imagine the ability to analyze single cells across multiple layers - their genes (transcriptomics), regulatory mechanisms (epigenomics), and proteins (proteomics) - all while pinpointing their exact location within the tissue! #SingleCell #SpatialTranscriptomics #Bioinformatics
Recent advancements in spatial omics have transformed genomics, allowing us to analyze cellular function within their native tissue environment. However, challenges like noise and limited resolution hinder current methods. Today's blog explores a novel technique that overcomes these limitations. Slide-Tag achieves high-throughput, single-nucleus barcoding with micrometer-scale spatial resolution and is compatible with various single-cell omics techniques. Read more: https://lnkd.in/dnmktd7w #SpatialTranscriptome #multiomics #singlecell #omics
-
-
⚡Last chance to submit your abstract for the 16th edition of 'Transcription and chromatin'! ⚡ You have only a week left to work on it so you can join this must-attend meeting brining together leading experts in the field and covering all aspects of transcription 👌🏻💯 ✍🏻 Submit your abstract by 3 June! Topics: 🧬 General basal transcriptional machinery 🧬 Pol II function, pausing and the role of RNA in transcription 🧬 Chromatin modifications and transcription 🧬 Chromatin readers, writers and epigenetic regulation 🧬 Condensates, hubs and speckles 🧬 Transcription regulation during embryonic development and disease 🧬 Genome topology and three-dimensional regulation 🧬 Quantitative and theoretical approaches 🧬 Imaging of transcription and genome 📅 24 – 27 August 2024 📍 EMBL Heidelberg and Virtual ➡️ s.embl.org/trm24-01 #EMBLTranscript
