RMH

Understanding protein architecture and predicting its structural tolerance to profound remodeling is pivotal for engineering functional proteins. We present SplitSeek-Pro, a deep learning model that evaluates amino acid-level splittability in folded proteins, a property critical for protein engineering tasks such as circular permutation and split reconstitution. By integrating primary sequences with 3D structural features, SplitSeek-Pro achieves residue-resolution predictions through a two-stage training process: large-scale pre-training followed by high-quality fine-tuning. Experimental validation on three distinct proteins confirms its superior predictive power over existing methods. Notably, SplitSeek-Pro identifies characteristic segments that function as cohesive, integral fragments analogous to super-secondary structural motifs. These results establish SplitSeek-Pro as a robust tool for rational protein engineering and offer insights into the fundamental structural building blocks of protein folding. To facilitate community access, we provide an automated web server at http://splitseek.topo.bio. Predicting protein splitability is pivotal for engineering functional variants. Here the authors present SplitSeek-Pro, a deep learning model integrating sequence and 3D features to achieve accurate residue-resolution split site prediction for protein design.

Accurate image annotation is essential for training artificial intelligence (AI) systems in biomedical image analysis, enabling tasks such as cell detection, tissue quantification, and disease characterization. However, creating pixel-level annotations is a time-consuming and labor-intensive process that requires expert input, limiting the development and adoption of AI methods. Recent advances in foundation models, such as the Segment Anything Model (SAM), enable interactive object segmentation from simple user prompts, but their integration into widely used bioimage analysis platforms remains limited and often requires technical expertise. Here we show that SAMJ, a user-friendly plugin for ImageJ/Fiji, enables fast, interactive, and accurate image annotation on standard computers without requiring programming skills or specialized hardware. SAMJ integrates efficient SAM variants into a familiar graphical interface, allowing users to delineate objects in large scientific images in real time using simple clicks or bounding boxes. This approach significantly reduces annotation effort, accelerates dataset creation, and broadens access to advanced AI-assisted annotation tools for the biomedical research community. García-López-de-Haro and colleagues present SAMJ, a plugin that brings the Segment Anything AI to ImageJ/Fiji, enabling fast, click-based image annotation on standard computers, and accelerating creation of biomedical training datasets.

Post-translational modifications are important for regulating cellular functions. Although traditional experimental methods accurately identify PTM sites, they are time-consuming. In this study, we propose a novel model capable of predicting 17 types of PTMs through multi-modal integration and AlphaFold predictions. Our model employs an enhanced CNN-transformer architecture to capture local dependencies within the sequence, while incorporating structural features and evolutionary patterns to effectively capture complex spatial relationships and global contextual dependencies. Through rigorous cross-validation and testing, our model demonstrates exceptional performance, achieving area under the curve scores of 96.5%, 91.6%, 91.0%, and 89.5% for the prediction of hydroxylation, malonylation, O-linked glycosylation, and phosphorylation, respectively. Notably, our model accurately identified known phosphorylation sites on tau and two recently identified residues linked to pre-tangle stages and early Alzheimer’s disease pathology. This work not only deepens the understanding of PTMs but also holds promise for advancing future research in the prediction of PTM sites and functional annotation.

Programs and computational frameworks for predicting RNA sequences with desired folding properties are continually being developed and expanded. A decade has passed since they were last reviewed in this journal, and this brief review provides an update to the review published at that time. Given a target secondary structure, these programs aim to predict RNA sequences that fold into the desired structure while satisfying various constraints. This procedure is known as inverse RNA folding. Traditionally, inverse RNA folding has been used to design optimized RNAs with favorable properties. This updated review covers some of the most widely used freeware programs developed for this purpose over the past decade. RNAinverse, part of the Vienna RNA package, was the first program devised to address the inverse RNA folding problem, and many subsequent programs were described in the earlier review. Some of the most important computational frameworks are the Infrared framework and DesiRNA. In addition, RNA design capabilities have been incorporated into the RNAstructure package, while NUPACK, as well as MoiRNAiFold, MODENA, incaRNAfbinv, and related tools have undergone recent updates. A variety of strategies have also emerged to address the problem of 3D RNA design and RNA–RNA interactions. The various programs mentioned employ distinct approaches, ranging from replica exchange Monte Carlo to constraint satisfaction, as well as Boltzmann sampling and machine learning approaches. Machine learning methods are being developed for emerging applications in biotechnology such as messenger RNA(mRNA) design and CRISPR guide RNA (gRNA) design. This brief review examines these programs and provides a timely update.

Acidobacteriota, one of the most abundant and ubiquitous bacterial phyla in soils, are well recognized for their role in carbon cycling. In contrast, their roles in soil nitrogen cycling remain largely unexplored, although recent metagenome-assembled genome (MAG) analyses suggest that Acidobacteriota may harbor genes involved in nitrogen cycling. Here, we provide culture-based evidence of diazotrophy within this phylum and demonstrate the widespread occurrence of nitrogen-fixing Acidobacteriota across diverse soil types. From grassland and agricultural soils, we isolated five Acidobacteriota strains representing novel taxonomic lineages, four of which harbor functional nitrogenase (nif) gene clusters. These strains were capable of fixing atmospheric nitrogen in vitro and/or in soil microcosms, as evidenced by acetylene reduction, N2-dependent growth, transcription of nif genes, incorporation of 15N into biomass and soil, and inhibition of nitrogenase activity by ammonium. Furthermore, global-scale meta-analysis of soil metagenomes revealed that nif-harboring Acidobacteriota are widely distributed and locally dominant across soil types. These results demonstrate the nitrogen-fixing capability of Acidobacteriota at the organismal level, complementing MAG-based inferences, and underscore their adaptive capacity in nitrogen-limited environments and their potential contribution to terrestrial nitrogen fixation. We also propose novel taxa within the class Terriglobia of the phylum Acidobacteriota, including diazotrophic strains, comprising one novel family, three novel genera, and four novel species: Koromonadaceae fam. nov., Koromonas soli gen. nov., sp. nov., Koromonas humicola sp. nov., Oryzophilus luti gen. nov., sp. nov., and Humiphilus diazotrophicus gen. nov., sp. nov.

Precise and orthogonal regulation of genetic circuits is a central challenge in synthetic biology, particularly at the translational level where tools remain scarce. Here, we address this by engineering suppressor transfer RNAs (sup-tRNAs) charged with canonical amino acids to enable programmable nonsense mutation readthrough in E. coli. Screening of 20 variants revealed a clear sup-tRNA design rule: readthrough efficiency is dictated by the similarity of the native tRNA anticodon to amber codon (CUA), as it preserves native aaRS interactions. We then demonstrate the power of this tool for advanced genetic circuit engineering. First, in a LacI-based biosensor, sup-tRNA regulation reduced background leakage by >77% and increased the induction dynamic range by 4.3-fold (from 6.67 to 28.68). Second, by dynamically balancing glycolytic flux through targeted pykA and pykF regulation, we increased the titer of N-acetylneuraminic acid by 66% (from 5.33 to 8.82 g/l) without compromising cell growth. Our work establishes engineered cAA-charged sup-tRNAs as a versatile, efficient, and cost-effective platform for precision translational control within genetic circuits, opening new avenues for biosensor optimization and metabolic engineering.