Bruce McKee

We have the pleasure of hosting Prof. Wen Ma from UVM Physics for seminar this Thursday (10:05 am, Innovation E105)! His talk will be on the "Unveiling Atomistic Mechanisms of Large Molecular Machines through the Computational Microscope." Abstract is below:

3

I wrote this news release for the Intelligent Computing journal article "Slimmed Optical Neural Networks with Multiplexed Neuron Sets and a Corresponding Backpropagation Training Algorithm." The article deals with a technique for optical computing, which is tricky but more energy efficient than using fully electronic computers. The authors combine wavelength-division multiplexing using their "multiplexed neuron sets" with a new backpropagation algorithm for machine learning to decrease the size of the optical network and thus its energy consumption compared to previous optical neural networks. Their work use semiconductor optical amplifiers but is applicable to other photonic devices as well. https://lnkd.in/gdAMug_b

4

Our latest work on MR image reconstruction

25

The "ChatGPT moment" for biology proceeds to unfold as protein language models, or machine learning tools trained on large databases of protein sequences, work to decode the language of biology with the goal of generating new proteins for widespread applications across therapeutics, sustainability, and more. Profluent has now announced proseLM, a new method which incorporates structural and functional context to protein language models for precise and steerable protein design. ProseLM was experimentally validated to be effective in improving gene editing activity and therapeutic antibody binding affinity, two complex functional protein design tasks. In protein design, structure-based methods and sequence-based approaches are "two sides of the same coin." While one side is “reading" biology by taking the sequence and determining the structure, the other is “writing" biology by generating a new protein that fits into a specific context. “[With proseLM], we can take the fine control of structure-based approaches and the broad scope of sequence-based approaches to look at the best of both worlds." - Jeffrey Ruffolo, Head of Protein Design at Profluent. Many thanks Ali Madani and Jeffrey Ruffolo for speaking with me! Read my full article on GEN Edge (Genetic Engineering & Biotechnology News): https://lnkd.in/gYXb7drA #AI #ProteinDesign #LanguageOfBiology #LLMs #FoundationalModel

44

2 Comments

Non-Linear Memory-Based Learning Advances Soil Property Prediction Using vis-NIR Spectral Data Researchers from Zhejiang University have developed a new non-linear memory-based learning (N-MBL) model that enhances the prediction accuracy of soil properties using visible near-infrared (vis-NIR) spectroscopy. By comparing N-MBL with traditional machine learning and local modeling methods, the study reveals its superior performance, particularly in predicting soil organic matter and total nitrogen. https://lnkd.in/gyGYD7Gx

3

Yesterday, David R. Liu published his lab's creation of #eePASSIGE, an "engineered and evolved" twist on prime editing that efficiently inserts large sequences of DNA into the genome. It's an important advance in the competitive gene insertion field. I covered the eePASSIGE data for Endpoints News when Liu first debuted it at ASGCT in May, along with gene insertion disclosures from Tome Bio, SalioGen, and Renegade (now owned by Orna). For more on those companies and the emerging field, check out my story: https://lnkd.in/gZFCW7CX #biotech #geneediting #CRISPR #genomeengineering #synbio #geneticmedicine

85

1 Comment

#ML revolutionizes de novo #PROTAC design! Researchers from Chalmers University of Technology shed light on current PROTAC development & ML's pivotal role, offering valuable insights for improved design strategies in this new modality. Quick Read: https://lnkd.in/g6RYCk7n #bioinformatics #drugdesign #machinelearning #ai #structuralbiology #sciencenews

5

We applaud David Baker (University of Washington), Demis Hassabis, and John Jumper (Google DeepMind) on receiving the 2024 The Nobel Prize in Chemistry for their pioneering work in computational #proteindesign and protein structure prediction. This recognition underscores how computational approaches are rapidly advancing biology and opening new frontiers in science. As we continue to push the boundaries of what’s possible, their contributions strengthen the entire field’s momentum toward a new era of discovery. #ProgrammingBiology | #GenerativeAI

185

1 Comment

Traditional visual systems are bulky, energy-intensive, and slow for cognitive tasks. Mimicking human visual systems shows promise, with #humanoid #MachineVision reducing background noise, adapting to lighting conditions, and improving sensing and processing. However, current systems lack polarimetric functionalities crucial for #navigation and #imaging. While interspecies-chimera vision could advance the field, integrating diverse visual functions into a single #optoelectronic #device remains challenging. Researchers from the University of Waterloo have developed an #optically-controlled polarimetry #memtransistor based on a van der Waals heterostructure, featuring polarization sensitivity, nonvolatility, and synaptic functions, enabling real-time navigation by identifying celestial polarizations like a honeybee. This innovation supports interspecies-chimera visual systems, benefiting #autonomous #vehicles, #medical #diagnoses, and intelligent #robotics. Now out 👉 #naturecommunications Nature Portfolio https://lnkd.in/ey-ywYyY

26

1 Comment

My latest Sci-Tech News post is live! Read about some cool news in science I found recently: https://lnkd.in/gsCx4e9H #scitech #science #tech #sciencenews

6

New post at Set Physics to Stun! Keeping stable funding for a scientific research group is stable, in academia or otherwise, is a challenging topic. Join SPTS in evaluating some ideas to approach this issue. https://lnkd.in/ejz8EdKE

8

This is worth a read. There is an urgent need to assess research outputs based on actual merit, rather than metics systems that can be corrupted. If you care about quality in science, scholarship and higher education, read this! #ResearchIntegrity #Fraud #Corruption #HigherEd #Academia #PublicationEthics

46

2 Comments

UC San Diego Center for Precision Computational Systems Network (PreCSN) is excited to launch its #monthly series--#FromOmics2OhMy!--which will feature examples of deriving #insights from #information (big data) by building #ActionableBiologicalNetworks. #Didyouknow that while #cancers are rampant in us (#humans), they #rare in the great #apes 🦍, with whom we share >99% #protein #sequence #homology! Why? A study ( 👇 link to article) that harnesses state-of-the-art #computational approaches at #PreCSN to bridge the gaps between #genetics, human #evolution, and #Siglec biology, reveals #insights with #significant #translational potential. By integrating #evolutionary #genetics, #phenome, and #transcriptome data, we uncovered how the evolutionary trajectory of the SIGLEC12 gene plays a crucial role in the transformation of epithelial cells and cancer development. But it wasn't until PreCSN's computational approaches were creatively applied to delve into the *nature* of this disparity, that the #insightful 💡 #link 🔗 between genomic alterations—particularly involving the SIGLEC12 gene—and ethnic disparity and inflammation in #colorectal #cancers emerged. The study demonstrated how Siglec-XII expression, which is present in about 30% of the human population due to a specific missense mutation, drives several pro-cancer phenotypes. On the other hand, the remaining 70% of humans, who have a frameshift mutation that eliminates the protein entirely, seem to be shielded from these effects. 🔬 Key highlights: Evolutionary #Insight💡: Our findings shed light on a previously unknown #genetic mechanism linking human evolution to cancer #susceptibility, particularly in #inflammation driven colorectal carcinomas. Siglec-XII's Role in Cancer: Demonstrated how SIGLEC12 expression in certain populations contributes to oncogenesis and affects prognosis. Cross-Species Comparison: This work underscores the importance of evolutionary genetics in understanding why certain cancers are uniquely prevalent in humans compared to our primate relatives. This study underscores the power of integrating computational tools with evolutionary biology to uncover the deep-rooted genetic factors that drive human disease, paving the way for targeted therapies and #precision #medicine. Link to article: https://lnkd.in/g7QFuQmH Shout out to the co-authors: Hector Cuello, Saptarshi Sinha, Andrea Verhagen; Nissi Varki, @avit varki, and Pradipta Ghosh, Center for Precision Computational Systems Network (PreCSN). Grateful to our funders: #NIAID, #NIH. Visit PreCSN: https://lnkd.in/gSfJwx3n #ContactUs: Saptarshi Sinha, #Director, #PreCSN [[email protected]] Post credit to: Saptarshi Sinha, #computationalbiologist, #physicist; #networkmedicine. #CancerResearch #ComputationalBiology #HumanEvolution #Genomics #SiglecBiology #Primates #PersonalizedMedicine #JCIInsight

9

Experts from NVIDIA, Stony Brook University, and Brookhaven National Laboratory have developed a charge-preserving algorithm (CPVQD) that streamlines excited state computations in quantum systems. 🌟 The algorithm focuses on reducing complexity with NVIDIA's CUDA-Q platform and the NERSC Perlmutter system for faster, more efficient results. https://lnkd.in/dKkZrEFV #Quantum #Algorithm #QuantumComputer #QuantumPlatform

23

Did you know that researchers from Lehigh University and Berkeley Lab have developed an Accelerating Sparse Accumulation (ASA) architecture, specialized computer hardware that enables faster computation on sparse matrices—datasets with many zero values? Their published work has led to a provisional patent, and the ASA architecture could accelerate a variety of valuable algorithms, including those impacting future metagenomic research. Learn more: https://bit.ly/LBNL_ASA ______________________________________________ cc: U.S. Department of Energy (DOE) U.S. Department of Energy Office of Science #datascience #science #computing #research #comptuerscience #HPC #computationalscience John Shalf Maximilian Bremer Cy Chan

20