Biology subjects

Cheng, J.

Publications and source records attributed to Cheng, J..

At least 73 records · Page 4Linked to original sources

bioRxiv2025

rRNA intermediates dictate nucleolar architecture in C. elegans

The nucleolus is a multilayered, membraneless organelle essential for rRNA transcription, processing and ribosome assembly; however, it remains largely unclear whether and how these molecular events coordinate the organization of the multilayered nucleolar structure. We previously reported that the accumulation of 27SA2 rRNA upon the loss of a particular subset of RPL proteins elicits nucleolar reshaping, resulting in the formation of nucleolar rings and vacuoles in C. elegans. Here, we systematically investigated the effects of pre-18S rRNA processing and rRNA transcription on nucleolar structure and reported that defects in these molecular events result in reorganization of the nucleolar architecture. In particular, the depletion of a special subset of small ribosomal subunit proteins (RPSs), namely, class I RPSs, induces a concentric three-layered nucleolar architecture; the inhibition of rRNA transcription induces another three-layered architecture, a sandwich-like nucleoli. Via fluorescence imaging, we found that nucleolar proteins undergo demixing and spatial redistribution into distinct subnucleolar regions during nucleolar structure reorganization. Tracking the subcellular localization of rDNA via a lacI::tagRFP/lacO system revealed that rDNA also redistributed upon nucleolar reorganization. We further identified the Ce.22S rRNA intermediate, which accumulates upon the loss of class I RPSs and regulates the reorganization of the nucleolar structure. Moreover, the conserved nucleolar protein NUCL-1/nucleolin is essential for nucleolar architecture reorganization and regulates the development of nematodes during impaired rRNA production. Together, these findings reveal that the nucleolar structure is orchestrated by the status of rRNA biogenesis and that the accumulation of particular rRNA intermediates can direct the organization of specific nucleolar architectures and the spatial mixing/demixing behavior of the nucleolar proteome.

genetics↗

bioRxiv2025

Intracranial Insights into the Developing Neural Basis of Moral Punishment

Third-party punishment (TPP) is a critical component of social regulation and justice, in-tegrating moral reasoning and emotional salience. However, the neural developmental basis underpinning this complex process remains largely unknown. Using rare intracranial stereo-electroencephalography (SEEG) in 14 children and 17 adults, we investigated the developing neural circuits of TPP. We found that broad-band gamma activity in the amygdala and ventromedial prefrontal cortex (vmPFC) encodes inferred intentions, with different patterns across age groups. Furthermore, the vmPFC, insula, and inferior parietal lobule (IPL) integrate punishment efficacy, also showing significant developmental differences. Combining task and resting state functional connectivity analyses, we further found age-dependent interactions among the amygdala-insula and IPL-vmPFC neural couplings during decision-making. These findings provide valuable intracranial evidence that the maturation of moral decision-making stems from the developmental refinement of subcortical-cortical circuits that integrate emotional and cognitive evaluations, explaining the shift from intuitive decisions in children to context-sensitive judgments in adults.

neuroscience↗

bioRxiv2025

HSD17B7 is required for Auditory Function by Regulating Cholesterol Synthesis in Sensory Hair Cells

Cholesterol homeostasis is fundamental to cellular function, and its disruption underlies a wide range of human diseases. However, the contribution of cholesterol biosynthesis to auditory physiology remains poorly understood. HSD17B7 (17{beta}-Hydroxysteroid dehydrogenase type 7) catalyzes the conversion of zymosterone to zymosterol, a key step in the post-lanosterol cholesterol biosynthetic pathway. Here, we found that Hsd17b7 is highly enriched in sensory hair cells of zebrafish and mice. The deficiency of Hsd17b7 deficiency reduced intracellular cholesterol levels in HEI-OC1 cells and zebrafish hair cells, thereby compromising MET and acoustic startle responses. A heterozygous nonsense variant (c.544G>T; p.E182*) in HSD17B7 was identified in an individual with bilateral profound hearing loss. mRNA of c.544G>T HSD17B7 failed to rescue the impaired MET and acoustic startle response of hsd17b7 mutants. Mechanistically, the mutation decreases mRNA abundance and significantly reduces protein. Moreover, expression of the p.E182* mutation disrupted the interaction between HSD17B7 and the ER retention receptor RER1, leading to aberrant subcellular localization and altered cholesterol distribution, thereby exacerbating HC dysfunction. Together, our findings suggest a conserved and essential role for HSD17B7-mediated cholesterol biosynthesis in sensory hair cell function and identify HSD17B7 as a candidate gene for sensorineural hearing loss.

developmental biology↗

bioRxiv2025

Multi-omics characterization of the skin microbiota reveals the anti-aging roles of Stenotrophomonas maltophilia

Shifts in the skin microbiome have shown a close link to chronological age. However, the contribution of skin microbiome in skin aging phenotypes remains unclear. To explore this, we performed phenotypic, metabolomic, metagenomic, and functional analyses on a cohort with divergent skin aging phenotypes. Genome-scale metabolic models (GEMs) integrated with metabolomic analysis revealed that Stenotrophomonas maltophilia, enriched in the younger group (categorized by AI-predicted age and skin elasticity), utilizes the glutathione cycle to maintain redox homeostasis. Cellular experiments showed its metabolites enhanced GSH synthesis and alleviated oxidative stress-induced skin aging by upregulating key genes in fibroblasts, including GCLM, PGD, SOD2, and NQO1. Additionally, GEMs highlighted its potential anti-aging roles in regulating host metabolic pathways involving betaine, lysolecithin, and porphyrin. In parallel, Acinetobacter guillouiae was found to influence host melanin metabolism by degrading dopamine (DA) and 3-methoxytyramine (3-MT), offering potential therapeutic strategies for mitigating pigmentation. Our findings highlight the dynamic interplay between skin microbiota and the host in aging, offering new insights for designing targeted anti-aging interventions.

microbiology↗

bioRxiv2025

Discovery of Niclosamide Analogs with Potent Mitochondrial Uncoupling Activity with reduced toxicity

Mitochondrial uncouplers have shown clinical potential across various diseases, including cancer. Niclosamide, an FDA-approved anthelmintic drug, acts as a mild mitochondrial uncoupler and has demonstrated anticancer activity in multiple preclinical cancer models. However, its clinical application remains limited, with some attributing this to poor bioavailability, while the underlying mechanisms are still unclear. Here, we demonstrate that niclosamide exhibits a dose-dependent biphasic effect, promoting uncoupling at low concentration while acting as a mitochondrial inhibitor at high concentration, which could restrict its therapeutic window and limit efficacy. To overcome this challenge, we aimed to develop next-generation mitochondrial uncouplers (MUs) by synthesizing and evaluating novel Niclosamide derivatives with enhanced therapeutic potential. Through structural modifications, we optimized uncoupling activity while reducing inhibitory toxicity, thereby expanding the pharmacological window. Our findings suggest that fine-tuning the molecular structure of mitochondrial uncouplers could provide a safer and more effective metabolic reprogramming strategy for cancer treatment.

pharmacology and toxicology↗

bioRxiv2025

Multimodal deep learning integration of cryo-EM and AlphaFold3 for high-accuracy protein structure determination

Cryo-electron microscopy (cryo-EM) is a key technology for determining the structures of proteins, particularly large protein complexes. However, automatically building high-accuracy protein structures from cryo-EM density maps remains a crucial challenge. In this work, we introduce MICA, a fully automatic and multimodal deep learning approach combining cryo-EM density maps with AlphaFold3-predicted structures at both input and output levels to improve cryo-EM protein structure modeling. It first uses a multi-task encoder-decoder architecture with a feature pyramid network to predict backbone atoms, C atoms and amino acid types from both cryo-EM maps and AlphaFold3-predicted structures, which are used to build an initial backbone model. This model is further refined using AlphaFold3-predicted structures and density maps to build final atomic structures. MICA significantly outperforms other state-of-the-art deep learning methods in terms of both modeling accuracy and completeness and is robust to protein size and map resolution. Additionally, it builds high-accuracy structural models with an average template-based modeling score (TM-score) of 0.93 from recently released high-resolution cryo-EM density maps, showing it can be used for real-world, automated, accurate protein structure determination.

bioinformatics↗

bioRxiv2025

AlphaGenome: advancing regulatory variant effect prediction with a unified DNA sequence model

Deep learning models that predict functional genomic measurements from DNA sequence are powerful tools for deciphering the genetic regulatory code. Existing methods trade off between input sequence length and prediction resolution, thereby limiting their modality scope and performance. We present AlphaGenome, which takes as input 1 megabase of DNA sequence and predicts thousands of functional genomic tracks up to single base pair resolution across diverse modalities - including gene expression, transcription initiation, chromatin accessibility, histone modifications, transcription factor binding, chro- matin contact maps, splice site usage, and splice junction coordinates and strength. Trained on human and mouse genomes, AlphaGenome matches or exceeds the strongest respective available external models on 24 out of 26 evaluations on variant effect prediction. AlphaGenomes ability to simultaneously score variant effects across all modalities accurately recapitulates the mechanisms of clinically-relevant variants near the TAL1 oncogene. To facilitate broader use, we provide tools for making genome track and variant effect predictions from sequence.

bioinformatics↗

bioRxiv2025

ERM proteins regulate the shape and number of Endoplasmic Reticulum-Plasma Membrane Junctions in neurons

Endoplasmic Reticulum (ER) - Plasma Membrane (PM) Junctions (EPJs) are specialized contact sites between ER membrane and the inner leaflet of PM. These junctions are critical for lipid exchange and Ca2+ signaling. In excitable cells like neurons and muscle, EPJs further modulate membrane excitability by regulating Ca2+ homeostasis. The mechanisms controlling EPJ abundance and morphology remain poorly understood. Using in vivo fluorescence imaging and electron microscopy of C. elegans neurons, we showed that EPJs form discrete, patch-like structures distributed across the soma. Through a forward genetics screen, we identified two conserved ERM (Ezrin-Radixin-Moesin) proteins, FRM-4 and FRM-1, as key regulators of EPJ shape and abundance. Both proteins localize to EPJs and exhibit liquid-liquid phase separation properties (LLPS). in vitro, purified FRM-4 binds to FRM-1, and together bundle filamentous actin. However, their presence in LLPS condensates and actin-bundling activity are mutually exclusive. In vivo, F-actin cables surround--but do not penetrate--EPJs, where FRM proteins are enriched as phase-separated condensates. Loss of FRM-4, FRM-1, or disruption of F-actin led to increased mobility of EPJs that fused into fewer but enlarged junctions. Together, our findings demonstrate that FRM-4 and FRM-1 control EPJ morphology by organizing peri-junctional F-actin networks, thereby restricting EPJ mobility and fusion.

cell biology↗

bioRxiv2025

Boosting AlphaFold Protein Tertiary Structure Prediction through MSA Engineering and Extensive Model Sampling and Ranking in CASP16

AlphaFold2 and AlphaFold3 have revolutionized protein structure prediction by enabling high-accuracy tertiary structure predictions for most single-chain proteins. However, obtaining high-quality predictions for some hard protein targets with shallow or noisy multiple sequence alignments (MSAs) and complicated multi-domain architectures remains challenging. Here, we present MULTICOM4, an integrative protein structure prediction system that uses diverse MSA generation, large-scale model sampling, and an ensemble model quality assessment (QA) strategy of combining individual QA methods to improve model generation and ranking of AlphaFold2 and AlphaFold3. In the 16th Critical Assessment of Techniques for Protein Structure Prediction (CASP16), our predictors built on MULTICOM4 ranked among the top performers out of 120 predictors in tertiary structure prediction and outperformed a standard AlphaFold3 predictor. The average TM-score of our best performing predictor MULTCOMs top-1 prediction for 84 CASP16 domain is 0.902. It achieved high accuracy (TM-score > 0.9) for 73.8% of the 84 domains and correct fold predictions (TM-score > 0.5) for 97.6% domains in terms of top-1 prediction. In terms of best-of-top-5 prediction, it predicted correct folds for all the domains. The results show that MSA engineering through the use of different protein sequence databases, alignment tools, and domain segmentation as well as extensive model sampling are the key to generate accurate and correct structural models. Additionally, using multiple complementary QA methods and model clustering can improve the robustness and reliability of model ranking.

bioinformatics↗

bioRxiv2025

A unified multimodal model for generalizable zero-shot and supervised protein function prediction

Predicting protein function is a fundamental yet challenging task that requires integrating diverse biological data modalities to capture complex functional relationships. Traditional machine learning methods often rely on single modalities or combine only a limited number (typically two), without aligning them in a unified representation, thereby constraining predictive accuracy. Moreover, most existing approaches are limited to preselected subsets of Gene Ontology (GO) function terms with sufficient annotations, making the prediction of novel function terms a persistent challenge. Here, we present FunBind, a multimodal AI model that jointly learns from protein sequences, textual descriptions, domain annotations, structural features, and GO terms to enhance prediction accuracy and infer previously unseen functions. FunBind operates in two modes: (1) self-supervised pretraining using contrastive learning to align the sequence modality with other heterogeneous modalities in a unified latent space, enabling unsupervised zero-shot function prediction, and (2) supervised fine-tuning that leverages all modalities for comprehensive and accurate function classification. Our results show that FunBinds zero-shot capabilities allow it to generalize effectively to novel function terms never encountered before, while its joint fine-tuning strategy substantially outperforms single-modality models and current state-of-the-art approaches in prediction accuracy.

bioinformatics↗

bioRxiv2025

Computational Assessment of Circularization Impact on Saxitoxin G-quadruplex Aptamer Performance

Although high-affinity and high-specificity aptamers can be obtained by SELEX (Systematic Evolution of Ligands by Exponential Enrichment), and have good activity in vitro, their performance in vivo is poor. This is mainly because of the degradation of linear nucleic acids by nucleases in body fluids and the difficulty of maintaining stable 3D structures in complex body fluid environments, which seriously hinders the clinical application of aptamers. This makes circular aptamers, which can resist degradation and are more stable in complex body fluid environments, an attractive option. However, the cyclization process may have a considerable impact on the structure of the original linear aptamer, and blind cyclization may lead to a decrease in its affinity. Here, we developed a new method to guide the cyclization of saxitoxin (STX) aptamers based on molecular computation, and verified the activity of circular aptamers through bioactivity experiments to further verify the authenticity and reliability of our methods. Consistent with the computational analysis, although circularization disrupts the G4 structure, circular aptamer not only has a higher affinity than the linear one but also provides certain protection to mice and greatly prolongs their survival, which proves the accuracy and reliability of the circular structure designed by our computational method. Our findings provide important ideas for better guidance in the design of circular aptamers and demonstrate their great potential in therapeutic applications.

biochemistry↗

bioRxiv2025

Developmentally-controlled generation of tRNA halves facilitates translational repression during sexual reproduction

The role of transfer RNAs (tRNAs) as mediators between the genetic code and protein synthesis is well established. In parallel, tRNAs can generate different types of functional small RNAs (sRNAs) that accumulate during stress and certain developmental processes across diverse organisms. Interestingly, this class of sRNAs (termed tRNA-derived sRNAs, tsRNAs) accumulate in the male gamete of mammals, insects, and plants but their role in these reproductive cells is unclear. Here, we determine the molecular pathway of tsRNA biogenesis in the plant male gamete-containing structure (the pollen grain) and identify their role in mediating the characteristic translational repression taking place in this tissue. Our data demonstrates that male-accumulating tsRNAs are generated by a cell-controlled pathway to aid in mediating reproductive translational repression and that their accumulation is key to ensuring the proper function of the pollen grain.

plant biology↗

bioRxiv2025

DICAR-JP Regulates Ribosome Migration, a New Theory of Mitochondrial Protein Production

BackgroundDICAR plays a cardioprotection of diabetic models. DICAR-JP, a short sequence of DICAR, may represent DICARs active functional domain. NAC is necessary for the heart tissue development. Addtionally, OGDHL is a metabolism regulator. In this study, we examine a new mechanism of DICAR via NAC/OGDHL on regulating cardiomyocyte metabolism in DCM. We also sought to elucidate the function of DICAR/DICAR-JP in modulating NAC-dependent production of OGDHL nascent peptides and their mitochondrial translocation, and we also aimed to optimize the DICAR-JP sequence to be better therapy on DCM. MethodsSPR was to investigate the binding interaction between DICAR-JP and NAC. The function of DICAR-JP and DICAR-JP45 on OGDHL nascent peptide transfection was detected by in vitro translation approach. Biotin-DICAR-JP was transfected to study the function of DICAR-JP, NAC and OGDHL. Untargeted metabolomics was utilized to characterize the metabolic reprogramming of cardiomyocytes. AAV9-DICAR-JP and DICAR-JP45 were constructed and administered to db/db mice for 1-2 months to assess their cardioprotective effects detected by Echocardiography confirmed mouse cardiac function. ResultsOur findings demonstrated that DICAR-JP be the functional domain that interacts with NAC to regulate OGDHL nascent peptide expression. Disruption of OGDHL expression reversed the metabolic reprogramming observed in diabetic cardiomyocytes, highlighting its crucial role in maintaining cardiac metabolic homeostasis. DICAR-JP facilitated the translocation of OGDHL nascent peptides from the cytoplasm to the mitochondria, leading us to hypothesize that DICAR-JP plays a key role in regulating ribosomal migration from the endoplasmic reticulum to the mitochondria. We refer to this process as the Ribosome Migration. In our studies using AAV9-mediated DICAR-JP and DICAR-JP45 overexpression in heart tissue. DICAR-JP and DICAR-JP45 both exhibited significant cardioprotective effects against diabetic cardiomyopathy (DCM), comparable to those observed with Empagliflozin. ConclusionsIn a conclusion, we propose a new endogenous nucleic acid candidate drug library, and a new molecular framework, the Ribosome Migration, which implicates the DICAR-JP/NAC/OGDHL nascent peptide axis in metabolic reprogramming related to DCM. This theory constructs a new mitochondrion protein from nuclear original protein. Moreover, DICAR-JP45 shows strong potential as a nucleic acid-based therapeutic candidate for treating DCM. Novelty and SignificanceO_ST_ABSWhat Is Known?C_ST_ABSDICAR is a new protective circular noncoding RNA for diabetic cardiomyopathy. NAC is located in ribosome and regulates new peptide production, which is also a key factor in heart tissue development. OGDHL can regulate different energy metabolism in mitochondria. What New Information Does This Article Contribute?DICAR-JP is the functional domain that interacts with NAC to regulate OGDHL nascent peptide expression. There is an interaction between DICAR-JP and NAC. DICAR-JP sequence is protected by NAC and NAC function is mediated by DICAR-JP. We called this RNA functional domain. OGDHL expression reversed the metabolic reprogramming observed in diabetic cardiomyocytes, highlighting its crucial role in maintaining cardiac metabolic homeostasis. DICAR-JP takes part in tanscription and translocation of OGDHL nascent peptides from the cytoplasm to the mitochondria via plasma ribosome. We call this process the Ribosome Migration Theory. DICAR-JP and DICAR-JP45 both exhibited significant cardioprotective effects against diabetic cardiomyopathy (DCM), comparable to those observed with Dapagliflozin (DAPA). Our study revealed a novel molecular framework, the Ribosome Migration Theory, which implicates the DICAR-JP/NAC/OGDHL nascent peptide axis in metabolic reprogramming related to DCM. Moreover, DICAR-JP45 shows strong potential as a nucleic acid-based therapeutic candidate for treating DCM.

pharmacology and toxicology↗

bioRxiv2025

MPBind: Multitask Protein Binding Site Prediction by Protein Language Models and Equivariant Graph Neural Networks

Proteins interact with a variety of molecules, including other proteins, DNAs, RNAs, ligands, ions, and lipids. These interactions play a crucial role in cellular communication, metabolic regulation, immune response, and structural integrity, making proteins fundamental to nearly all biological functions. Accurately predicting protein binding (interaction) sites is essential for understanding protein interactions and functions. Here, we introduce MPBind, a multitask protein binding site prediction method, which integrates protein language models (PLMs) that can extract structural and functional information from sequences and equivariant graph neural networks (EGNNs) that can effectively capture geometric features of 3D protein structures. Through multitask learning, it can predict binding sites on proteins that interact with five key categories of binding partners: proteins, DNA/RNA, ligands, lipids, and ions, which is more comprehensive than most existing task-specific methods that can predict only one or a few kinds of binding sites. Moreover, MPBind outperforms both general bind site prediction methods and task-specific binding site prediction methods.

bioinformatics↗

bioRxiv2025

The H3K27me3 reader UAD-2 recruits a TAF-12-containing transcription condensates to initiate piRNA expression within heterochromatic clusters

Metazoans utilize the small RNA pathway to regulate gene expression and maintain genome integrity. This pathway directs histone H3 lysine 9 tri-methylation (H3K9me3) or histone H3 lysine 27 tri-methylation (H3K27me3) at target loci to induce transcriptional gene silencing. Interestingly, some small RNAs are generated from genomic loci enriched in H3K9me3 or H3K27me3. However, the transcription mechanism of small RNA precursors from these heterochromatic regions remains unclear. In C. elegans, piRNAs originate from two genomic clusters enriched with H3K27me3 marks, which recruit the H3K27me3 reader UAD-2 and the upstream sequence transcription complex (USTC). Here, we demonstrate that piRNA transcription in C. elegans relies on TAF-12, a subunit of the basal transcription factor IID (TFIID). Depletion of TAF-12 reduces the production of both piRNA precursors and mature piRNAs. TAF-12 interacts with UAD-2 and facilitates piRNA focus formation in germ cell nuclei. We further show that TAF-12 triggers piRNA transcription by recruiting the RNA polymerase II subunit RPB-5, the Mediator complex subunit MDT-8, and the general transcription factors GTF-2F2 and GTF-2H2C. Thus, piRNA transcription within heterochromatic regions depends on the collaboration between histone modification readers, piRNA-specific transcription factors, and core transcription machinery.

genetics↗

bioRxiv2025

A Labeled Dataset for AI-based Cryo-EM Map Enhancement

Cryo-electron microscopy (cryo-EM) has transformed structural biology by enabling near-atomic resolution imaging of macromolecular complexes. However, cryo-EM density maps suffer from intrinsic noise arising from structural sources, shot noise, and digital recording, which complicates accurate atomic structure building. While various methods for denoising cryo-EM density maps exist, there is a lack of standardized datasets for benchmarking artificial intelligence (AI) approaches. Here, we present an open-source dataset for cryo-EM density map denoising comprising 650 high-resolution (1-4 [A]) experimental maps paired with three types of generated label maps: regression maps capturing idealized density distributions, binary classification maps distinguishing structural elements from background, and atom-type classification maps. Each map is standardized to 1 [A] voxel size and validated through Fourier Shell Correlation analysis, demonstrating substantial resolution improvements in label maps compared to experimental maps. This resource bridges the gap between structural biology and artificial intelligence communities, enabling researchers to develop and benchmark innovative methods for enhancing cryo-EM density maps.

bioinformatics↗

bioRxiv2025

Atomic Protein Structure Modeling from Cryo-EM Using Multi-Modal Deep Learning and AlphaFold3

Cryo-electron microscopy (cryo-EM) has revolutionized structural biology by enabling near-atomic resolution visualization of protein structures. However, accurately modeling 3D atomic structures from cryo-EM density maps remains challenging, particularly for large multi-chain complexes and assemblies. Here, we present an automated pipeline that integrates multi-modal deep learning and advanced structure prediction techniques to improve model accuracy. Our approach leverages a deep learning model that combines sequence-based features from a Protein Language Model with cryo-EM density maps, enabling a richer feature representation across modalities. The deep learning-predicted voxels are utilized to build a Hidden Markov Model (HMM) and a tailored Viterbi algorithm is used to align sequences to generate an initial protein backbone structures. These backbone models then serve as templates for AlphaFold3, which refines the structures for improved accuracy. Our approach combines cryo-EM data with AlphaFold3 predictions, helping to refine and improve AlphaFold3s predicted structures. By integrating both methods, we can generate more accurate and reliable atomic models, particularly for large proteins with complex conformations.

bioinformatics↗

bioRxiv2025

Improving AlphaFold2 and 3-based protein complex structure prediction with MULTICOM4 in CASP16

With AlphaFold achieving high-accuracy tertiary structure prediction for most single-chain proteins (monomers), the next major challenge in protein structure prediction is accurately modeling multi-chain protein complexes (multimers). We developed MULTICOM4, the latest version of the MULTICOM system, to improve protein complex structure prediction by integrating transformer-based AlphaFold2, diffusion model-based AlphaFold3, and our in-house techniques. These include protein complex stoichiometry prediction, diverse multiple sequence alignment (MSA) generation leveraging both sequence and structure comparison, modeling exception handling, and deep learning-based model quality assessment. MULTICOM4was blindly evaluated in the 16th communitywide Critical Assessment of Techniques for Protein Structure Prediction (CASP16) in 2024. In Phase 0 of CASP16, where stoichiometry information was unavailable, MULTICOM predictors performed best, with MULTICOM_human achieving a TM-score of 0.752 and a DockQ score of 0.584 for top-ranked predictions on average. In Phase 1 of CASP16, with stoichiometry information provided, MULTICOM_human remained among the top predictors, attaining a TM-score of 0.797 and a DockQ score of 0.558 on average. The CASP16 results demonstrate that integrating complementary AlphaFold2 and 3 with enhanced MSA inputs, comprehensive model ranking, exception handling, and accurate stoichiometry prediction can effectively improve protein complex structure prediction.

bioinformatics↗