搜索结果：Molecular biomedicine

FIR and submm observations have established the fundamental role of dust-obscured star formation in the assembly of stellar mass over the past 12 billion years. At z between 2 and 4, the bulk of star formation is enshrouded in dust, and dusty star forming galaxies (DSFGs) contain about half of the total stellar mass density. Star formation develops in dense molecular clouds, and is regulated by a complex interplay between all the ISM components that contribute to the energy budget of a galaxy: gas, dust, cosmic rays, interstellar electromagnetic fields, gravitational field, dark matter. Molecular gas is the actual link between star forming gas and its complex environment, providing by far the richest amount of information about the star formation process. However, molecular lines interpretation requires complex modeling of astrochemical networks, which regulate the molecular formation and establishes molecular abundances in a cloud, and a modeling of the physical conditions of the gas in which molecular energy levels become populated. This paper critically reviews the main astrochemical parameters needed to get predictions about molecular signals in DSFGs. We review the current kno

Molecular systems involve interactions across multiple spatial scales, from local coordination and short-range perturbations to long-range electrostatic and solvent-mediated effects. However, most molecular representation learning methods rely on manually predefined scales, and the task-optimal modeling scale may not coincide with these fixed levels. This study introduces a loss-guided adaptive scale refinement framework for molecular force prediction, treating predefined scales as initial anchors and discovering task-effective resolutions through interpolation, routing, differentiable scale updates, and scale pool refinement. Using a NaCl aqueous ionic system as a minimal testbed, this study constructs short-scale and long-range force prediction branches and analyzes their complementarity. Oracle hard routing reduces the overall force MAE from 399.65 to 382.67, while continuous oracle interpolation further reduces it to 380.96. In close-contact regimes with nearest-ion distance below 0.6 nm, the close-contact MAE decreases from 327.22 to 260.51. A minimal scale pool update experiment shows that starting from endpoint anchors {0,1}, loss-guided updates automatically generate interm

AI-assisted molecular property prediction has become a promising technique in early-stage drug discovery and materials design in recent years. However, due to high-cost and complex wet-lab experiments, real-world molecules usually experience the issue of scarce annotations, leading to limited labeled data for effective supervised AI model learning. In light of this, few-shot molecular property prediction (FSMPP) has emerged as an expressive paradigm that enables learning from only a few labeled examples. Despite rapidly growing attention, existing FSMPP studies remain fragmented, without a coherent framework to capture methodological advances and domain-specific challenges. In this work, we present the first comprehensive and systematic survey of few-shot molecular property prediction. We begin by analyzing the few-shot phenomenon in molecular datasets and highlighting two core challenges: (1) cross-property generalization under distribution shifts, where each task corresponding to each property, may follow a different data distribution or even be inherently weakly related to others from a biochemical perspective, requiring the model to transfer knowledge across heterogeneous predi

Molecular communication (MC) provides a foundational framework for information transmission in the Internet of Bio-Nano Things (IoBNT), where efficiency and reliability are crucial. However, the inherent limitations of molecular channels, such as low transmission rates, noise, and intersymbol interference (ISI), limit their ability to support complex data transmission. This paper proposes an end-to-end semantic learning framework designed to optimize task-oriented molecular communication, with a focus on biomedical diagnostic tasks under resource-constrained conditions. The proposed framework employs a deep encoder-decoder architecture to efficiently extract, quantize, and decode semantic features, prioritizing taskrelevant semantic information to enhance diagnostic classification performance. Additionally, a probabilistic channel network is introduced to approximate molecular propagation dynamics, enabling gradient-based optimization for end-to-end learning. Experimental results demonstrate that the proposed semantic framework improves diagnostic accuracy by at least 25% compared to conventional JPEG compression with LDPC coding methods under resource-constrained communication sce

Recent development in Retrieval-Augmented Large Language Models (LLMs) have shown great promise in biomedical applications. How ever, a critical gap persists in reliably evaluating their curation ability the process by which models select and integrate relevant references while filtering out noise. To address this, we introduce the benchmark for Curation of Retrieval-Augmented LLMs in Biomedicine (CRAB), the first multilingual benchmark tailored for evaluating the biomedical curation of retrieval-augmented LLMs, available in English, French, German and Chinese. By incorporating a novel citation-based evaluation metric, CRAB quantifies the curation performance of retrieval-augmented LLMs in biomedicine. Experimental results reveal significant discrepancies in the curation performance of mainstream LLMs, underscoring the urgent need to improve it in the domain of biomedicine. Our dataset is available at https://huggingface.co/datasets/zhm0/CRAB.

Information molecules play a crucial role in molecular communication (MC), acting as carriers for information transfer. A common approach to get information molecules in MC involves harvesting them from the environment; however, the harvested molecules are often a mixture of various environmental molecules, and the initial concentration ratios in the reservoirs are identical, which hampers high-fidelity transmission techniques such as molecular shift keying (MoSK). This paper presents a transmitter design that harvests molecules from the surrounding environment and stores them in two reservoirs. To separate the mixed molecules, energy is consumed to transfer them between reservoirs. Given limited energy resources, this work explores energy-efficient strategies to optimize transmitter performance. Through theoretical analysis and simulations, we investigate different methods for moving molecules between reservoirs. The results demonstrate that transferring higher initial concentration molecules enhances transmitter performance, while using fewer molecules per transfer further improves efficiency. These findings provide valuable insights for optimizing MC systems through energy-effic

Recent breakthroughs in large language models (LLMs) offer unprecedented natural language understanding and generation capabilities. However, existing surveys on LLMs in biomedicine often focus on specific applications or model architectures, lacking a comprehensive analysis that integrates the latest advancements across various biomedical domains. This review, based on an analysis of 484 publications sourced from databases including PubMed, Web of Science, and arXiv, provides an in-depth examination of the current landscape, applications, challenges, and prospects of LLMs in biomedicine, distinguishing itself by focusing on the practical implications of these models in real-world biomedical contexts. Firstly, we explore the capabilities of LLMs in zero-shot learning across a broad spectrum of biomedical tasks, including diagnostic assistance, drug discovery, and personalized medicine, among others, with insights drawn from 137 key studies. Then, we discuss adaptation strategies of LLMs, including fine-tuning methods for both uni-modal and multi-modal LLMs to enhance their performance in specialized biomedical contexts where zero-shot fails to achieve, such as medical question answ

Research resources (RRs) such as data, software, and tools are essential pillars of scientific research. The field of biomedicine, a critical scientific discipline, is witnessing a surge in research publications resulting in the accumulation of a substantial number of RRs. However, these resources are dispersed among various biomedical articles and can be challenging to locate and reuse due to their transient nature. In this paper, we report our recent progress in biomedical data curation - building a large research resource database for biomedicine (RRD-Bio), based on a collection of 40 million papers from two large biomedical literature databases, PubMed and PubMed Central. The database contains 2,555,116 RRs, each identified by a location on the Internet (URL) and descriptive information (Context). We made the RRD-Bio database publicly available (\url{https://zenodo.org/records/10526493}) to enhance the visibility of biomedical research resources, the ability to preserve important resources and the reproducibility of biomedical research.

Large Language Models (LLMs) have demonstrated remarkable capabilities across various domains and are moving towards more specialized areas. Recent advanced proprietary models such as GPT-4 and Gemini have achieved significant advancements in biomedicine, which have also raised privacy and security challenges. The construction of specialized generalists hinges largely on high-quality datasets, enhanced by techniques like supervised fine-tuning and reinforcement learning from human or AI feedback, and direct preference optimization. However, these leading technologies (e.g., preference learning) are still significantly limited in the open source community due to the scarcity of specialized data. In this paper, we present the UltraMedical collections, which consist of high-quality manual and synthetic datasets in the biomedicine domain, featuring preference annotations across multiple advanced LLMs. By utilizing these datasets, we fine-tune a suite of specialized medical models based on Llama-3 series, demonstrating breathtaking capabilities across various medical benchmarks. Moreover, we develop powerful reward models skilled in biomedical and general reward benchmark, enhancing fur

ChatGPT has drawn considerable attention from both the general public and domain experts with its remarkable text generation capabilities. This has subsequently led to the emergence of diverse applications in the field of biomedicine and health. In this work, we examine the diverse applications of large language models (LLMs), such as ChatGPT, in biomedicine and health. Specifically we explore the areas of biomedical information retrieval, question answering, medical text summarization, information extraction, and medical education, and investigate whether LLMs possess the transformative power to revolutionize these tasks or whether the distinct complexities of biomedical domain presents unique challenges. Following an extensive literature survey, we find that significant advances have been made in the field of text generation tasks, surpassing the previous state-of-the-art methods. For other applications, the advances have been modest. Overall, LLMs have not yet revolutionized biomedicine, but recent rapid progress indicates that such methods hold great potential to provide valuable means for accelerating discovery and improving health. We also find that the use of LLMs, like Chat

The function of the organism hinges on the performance of its information-processing networks, which convey information via molecular recognition. Many paths within these networks utilize molecular codebooks, such as the genetic code, to translate information written in one class of molecules into another molecular "language" . The present paper examines the emergence and evolution of molecular codes in terms of rate-distortion theory and reviews recent results of this approach. We discuss how the biological problem of maximizing the fitness of an organism by optimizing its molecular coding machinery is equivalent to the communication engineering problem of designing an optimal information channel. The fitness of a molecular code takes into account the interplay between the quality of the channel and the cost of resources which the organism needs to invest in its construction and maintenance. We analyze the dynamics of a population of organisms that compete according to the fitness of their codes. The model suggests a generic mechanism for the emergence of molecular codes as a phase transition in an information channel. This mechanism is put into biological context and demonstrated

The estimation of molecular abundances in interstellar clouds from spectroscopic observations requires radiative transfer calculations, which depend on basic molecular input data. This paper reviews recent developments in the fields of molecular data and radiative transfer. The first part is an overview of radiative transfer techniques, along with a "road map" showing which technique should be used in which situation. The second part is a review of measurements and calculations of molecular spectroscopic and collisional data, with a summary of recent collisional calculations and suggested modeling strategies if collision data are unavailable. The paper concludes with an overview of future developments and needs in the areas of radiative transfer and molecular data.

This contribution exploits the duality between a viral infection process and macroscopic air-based molecular communication. Airborne aerosol and droplet transmission through human respiratory processes is modeled as an instance of a multiuser molecular communication scenario employing respiratory-event-driven molecular variable-concentration shift keying. Modeling is aided by experiments that are motivated by a macroscopic air-based molecular communication testbed. In artificially induced coughs, a saturated aqueous solution containing a fluorescent dye mixed with saliva is released by an adult test person. The emitted particles are made visible by means of optical detection exploiting the fluorescent dye. The number of particles recorded is significantly higher in test series without mouth and nose protection than in those with a wellfitting medical mask. A simulation tool for macroscopic molecular communication processes is extended and used for estimating the transmission of infectious aerosols in different environments. Towards this goal, parameters obtained through self experiments are taken. The work is inspired by the recent outbreak of the coronavirus pandemic.

Existing molecular communication systems, both theoretical and experimental, are characterized by low information rates. In this paper, inspired by time-of-flight mass spectrometry (TOFMS), we consider the design of a molecular communication system in which the channel is a vacuum and demonstrate that this method has the potential to increase achievable information rates by many orders of magnitude. We use modelling results from TOFMS to obtain arrival time distributions for accelerated ions and use them to analyze several species of ions, including hydrogen, nitrogen, argon, and benzene. We show that the achievable information rates can be increased using a velocity (Wien) filter, which reduces uncertainty in the velocity of the ions. Using a simplified communication model, we show that data rates well above 1 Gbit/s/molecule are achievable.

Molecular recognition, which is essential in processing information in biological systems, takes place in a crowded noisy biochemical environment and requires the recognition of a specific target within a background of various similar competing molecules. We consider molecular recognition as a transmission of information via a noisy channel and use this analogy to gain insights on the optimal, or fittest, molecular recognizer. We focus on the optimal structural properties of the molecules such as flexibility and conformation. We show that conformational changes upon binding, which often occur during molecular recognition, may optimize the detection performance of the recognizer. We thus suggest a generic design principle termed 'conformational proofreading' in which deformation enhances detection. We evaluate the optimal flexibility of the molecular recognizer, which is analogous to the stochasticity in a decision unit. In some scenarios, a flexible recognizer, i.e., a stochastic decision unit, performs better than a rigid, deterministic one. As a biological example, we discuss conformational changes during homologous recombination, the process of genetic exchange between two DNA s

The CDMS was founded 1998 to provide in its catalog section line lists of molecular species which may be observed in various astronomical sources using radio astronomy. The line lists contain transition frequencies with qualified accuracies, intensities, quantum numbers, as well as further auxilary information. They have been generated from critically evaluated experimental line lists, mostly from laboratory experiments, employing established Hamiltonian models. Seperate entries exist for different isotopic species and usually also for different vibrational states. As of December 2015, the number of entries is 792. They are available online as ascii tables with additional files documenting information on the entries. The Virtual Atomic and Molecular Data Centre was founded more than 5 years ago as a common platform for atomic and molecular data. This platform facilitates exchange not only between spectroscopic databases related to astrophysics or astrochemistry, but also with collisional and kinetic databases. A dedicated infrastructure was developed to provide a common data format in the various databases enabling queries to a large variety of databases on atomic and molecular dat

Molecular codes translate information written in one type of molecules into another molecular language. We introduce a simple model that treats molecular codes as noisy information channels. An optimal code is a channel that conveys information accurately and efficiently while keeping down the impact of errors. The equipoise of the three conflicting needs, for minimal error-load, minimal cost of resources and maximal diversity of vocabulary, defines the fitness of the code. The model suggests a mechanism for the emergence of a code when evolution varies the parameters that control this equipoise and the mapping between the two molecular languages becomes non-random. This mechanism is demonstrated by a simple toy model that is formally equivalent to a mean-field Ising magnet.

Cryo-electron microscopy (cryo-EM) has become a major experimental technique to determine the structures of large protein complexes and molecular assemblies, as evidenced by the 2017 Nobel Prize. Although cryo-EM has been drastically improved to generate high-resolution three-dimensional (3D) maps that contain detailed structural information about macromolecules, the computational methods for using the data to automatically build structure models are lagging far behind. The traditional cryo-EM model building approach is template-based homology modeling. Manual de novo modeling is very time-consuming when no template model is found in the database. In recent years, de novo cryo-EM modeling using machine learning (ML) and deep learning (DL) has ranked among the top-performing methods in macromolecular structure modeling. Deep-learning-based de novo cryo-EM modeling is an important application of artificial intelligence, with impressive results and great potential for the next generation of molecular biomedicine. Accordingly, we systematically review the representative ML/DL-based de novo cryo-EM modeling methods. And their significances are discussed from both practical and methodolo

G-Protein Coupled Receptors (GPCRs) are a big family of eukaryotic cell transmembrane proteins, responsible for numerous biological processes. From a practical viewpoint around 34\% of the drugs approved by the US Food and Drug Administration target these receptors. They can be analyzed from their simulated molecular dynamics, including the prediction of their behavior in the presence of drugs. In this paper, the capability of Long Short-Term Memory Networks (LSTMs) are evaluated to learn and predict the molecular dynamic trajectories of a receptor. Several models were trained with the 3D position of the amino acids of the receptor considering different transformations on the position of the amino acid, such as their centers of mass, the geometric centers and the position of the $α$--carbon for each amino acid. The error of the prediction of the position was evaluated by the mean average error (MAE) and root-mean-square deviation (RMSD). The LSTM models show a robust performance, with results comparable to the state-of-the-art in non-dynamic 3D predictions. The best MAE and RMSD values were found for the mass center of the amino acids with 0.078 Å and 0.156 Å respectively. This wor