Search for: All records

Abstract Viruses are key drivers of microbial ecology and evolution, yet their study is hindered due to challenges in culturing. Traditional gene-centric methods, which focus on a few hallmark genes like for capsids, miss much of the viral genome, leaving key viral proteins and functions undiscovered. Here, we introduce two powerful annotation-free metrics, V-score and V_L-score, designed to quantify the “virus-likeness” of protein families and genomes and create an open-access searchable database, ‘V-Score-Search’. By applying V- and V_L-scores to public protein databases, we link 19 − 59% of protein families with viruses representing a 5 − 8x increase over current estimates. These metrics outperform existing approaches, enabling high efficiency in detection of viral genomes, prophages, and host-derived auxiliary viral genes (AVGs) from fragmented sequences. Remarkably, we identify up to 17 times more AVGs dominated by non-metabolic proteins of unknown function. This innovation unlocks new insights into virus signatures and host interactions, with wide-ranging implications from genomics to biotechnology. 

Many critical information technology and cyber-physical systems rely on a supply chain of open-source software projects. OSS project maintainers often integrate contributions from external actors. While maintainers can assess the correctness of a pull request, assessing a pull request’s cybersecurity implications is challenging. To help maintainers make this decision, we propose that the open-source ecosystem should incorporate Actor Reputation Metric Systems (ARMS). This capability would enable OSS maintainers to assess a prospective contributor’s cybersecurity reputation. To support the future instantiation of ARMS, we identify seven generic security signals from industry standards; map concrete metrics from prior work and available security tools; describe study designs to refine and assess the utility of ARMS; and finally weigh its pros and cons. 

Deep learning has achieved recognition for its impact within natural sciences, yet the prohibitive financial and technical cost of training models from scratch inhibits adoption. Following software engineering community guidance, natural scientists are reusing pre-trained deep learning models (PTMs) to amortize these costs. While prior works recommend PTM reuse patterns, we present the first empirical study of these patterns in the natural sciences, quantifying the utilization and impact of PTM reuse within the scientific process across 17,718 peer-reviewed, open-access papers. Our results show that “Biochemistry, Genetics and Molecular Biology” has outpaced other natural scientific fields in PTM reuse, “adaptation” reuse is the most prevalent PTM reuse pattern identified across all natural science fields, and the “testing” stage of the scientific process has been most impacted by PTM integration. 

Automatic software system optimization can improve software speed, reduce operating costs, and save energy. Traditional approaches to optimization rely on manual tuning and compiler heuristics, limiting their ability to generalize across diverse codebases and system contexts. Recent methods using Large Language Models (LLMs) introduce automation on simple programs, but they do not scale effectively to the complexity and size of real-world software systems. We present SysLLMatic, a system that integrates LLMs with performance diagnostics and a curated catalog of 43 optimization patterns to automatically optimize software systems. By leveraging profiling to identify performance hotspots, our approach enables LLMs to optimize real-world software beyond isolated code snippets. We evaluate it on three benchmark suites: HumanEval_CPP (competitive programming in C++), SciMark2 (scientific kernels in Java), and DaCapo (large-scale software systems in Java). Results show that SysLLMatic can improve software system performance, including latency, throughput, energy efficiency, memory usage, and CPU utilization. It consistently outperforms state-of-the-art LLM baselines on microbenchmarks. On largescale application codes, to which prior LLM approaches have not scaled, it surpasses compiler optimizations, achieving average relative improvements of 1.54× in latency (vs. 1.01× for the compiler) and 1.24× in energy (vs. 1.08× for the compiler). Our findings demonstrate that LLMs, guided by performance knowledge through the optimization pattern catalog and appropriate performance diagnostics, can serve as viable software system optimizers. We further identify limitations of our approach and the challenges involved in handling complex applications. This work provides a foundation for generating optimized code across various languages, benchmarks, and program sizes in a principled manner. 

The species historically referred to as Buellia stillingiana, and more recently proposed to be a synonym of B. erubescens, is very common and widespread in eastern North America. While studying material to determine the application of these names, we repeatedly encountered a lichenicolous coelomycete parasitizing thalli of this lichen. The lichenicolous fungus appears to belong to the genus Minutophoma, differing from M. chrysophthalmae in its larger pycnidia, larger conidia, and occurrence on Buellia rather than Chrysothrix. It is here described as M. buelliarum. After studying more than one thousand specimens of Buellia stillingiana, M. buelliarum was found, albeit rarely, throughout the range of this lichen from the island of Newfoundland in eastern Canada south throughout the Coastal Plain and west to the edge of the Great Plains. It was also found to occur on other morphologically similar sympatric Buellia species producing norstictic acid: B. curtisii, B. imshaugiana, and an unidentified Buellia species similar to B. stillingiana. 

Large language models and AI agents have recently shown promise in automating software performance optimization, but existing approaches predominantly rely on local, syntax-driven code transformations. This limits their ability to reason about program behavior and capture whole system performance interactions. As modern software increasingly comprises interacting components — such as microservices, databases, and shared infrastructure — effective code optimization requires reasoning about program structure and system architecture beyond individual functions or files. This paper explores the feasibility of whole system optimization for microservices. We introduce a multi-agent framework that integrates control-flow and data-flow representations with architectural and cross-component dependency signals to support system-level performance reasoning. The proposed system is decomposed into coordinated agent roles—summarization, analysis, optimization, and verification—that collaboratively identify crosscutting bottlenecks and construct multi-step optimization strategies spanning the software stack. We present a proof-of-concept on a microservice-based system that illustrates the effectiveness of our proposed framework, achieving a 36.58% improvement in throughput and a 27.81% reduction in average response time. 

Modern transformer architectures achieve remarkable performance across tasks and domains but remain rigid in how they allocate computation at inference time. Real-world deployment often requires models to adapt to diverse hardware and latency constraints, yet most approaches to dynamic computation focus on a single axis --- such as reducing the number of tokens.We present a novel capability: AdaPerceiver, the first transformer architecture with unified adaptivity across depth, width, and tokens within a single model.We propose an architecture that supports adaptivity along these axes.We couple this with an efficient joint training regime that ensures the model maintains performance across its various configurations.We evaluate AdaPerceiver on image classification, semantic segmentation, and depth estimation tasks.On image classification, AdaPerceiver expands the accuracy-throughput Pareto front.It achieves 85.4% accuracy while yielding 36% higher throughput than FlexiViT-L.On dense prediction, AdaPerceiver matches ViT-H14 while having ~26x fewer encoder FLOPs (floating-point operations) on semantic segmentation and depth estimation.Finally, we show how AdaPerceiver equipped with a policy can maintain ImageNet1K accuracy (0.1 percentage points) while reducing FLOPs by 24-33%. 

Diffusion models are state-of-the-art generative models, yet their samples often fail to satisfy application objectives such as safety constraints or domain-specific validity. Existing techniques for alignment require gradients, internal model access, or large computational budgets --- resulting in high compute demands, or lack of support for certain objectives. In response, we introduce an inference-time alignment framework based on evolutionary algorithms. We treat diffusion models as black-boxes and search their latent space to maximize alignment objectives. Given equal or less running time, our method achieves 3-35% higher ImageReward scores than gradient-free and gradient-based methods. On the Open Image Preferences dataset, our methods achieve competitive results across four popular alignment objectives. In terms of computational efficiency, we require 55% to 76% less GPU memory and are 72% to 80% faster than gradient-based methods. 

Empirical studies of research software are hard to compare because the literature operationalizes “research software” inconsistently. Motivated by the research software supply chain (RSSC) and its security risks, we introduce an RSSC-oriented taxonomy that makes scope and operational boundaries explicit for empirical research software security studies. We conduct a targeted scoping review of recent repository mining and dataset construction studies, extracting each work’s definition, inclusion criteria, unit of analysis, and identification heuristics. We synthesize these into a harmonized taxonomy and a mapping that translates prior approaches into shared taxonomy dimensions. We operationalize the taxonomy on a large community-curated corpus from the Research Software Encyclopedia (RSE), producing an annotated dataset, a labeling codebook, and a reproducible labeling pipeline. Finally, we apply OpenSSF Scorecard as a preliminary security analysis to show how repository-centric security signals differ across taxonomy-defined clusters and why taxonomy-aware stratification is necessary for interpreting RSSC security measurements.