Three-dimensional (3D) genomic architecture is crucial for the regulation of different biological processes, and the study of multi-way chromatin structures represents a cutting-edge area in this field. Pore-C is an advanced experimental technique integrating chromosome conformation capture (3C) with Nanopore long-read sequencing, which can effectively capture complex multi-way chromatin contacts across the genome. Due to the absence of comprehensive and dedicated tools, the analysis of Pore-C data remains challenging, restricting its broader application. To address this limitation, we developed Pore-C Pipeline Toolbox (PPL-Toolbox), a specialized software for a comprehensive analysis of multi-way chromatin interactions from Pore-C data. PPL-Toolbox incorporates a suite of optimized modules, including quality control, multi-way interaction extraction, noise reduction for multi-way contacts, advanced visualization tools for multi-way contacts, and haplotype map generation. Evaluation on both simulated and real datasets demonstrates that the PPL-Toolbox outperforms existing methods by providing a more comprehensive set of results. PPL-Toolbox is expected to become a powerful and versatile tool, driving advance in multi-way 3D genomics research, and enabling new discoveries in the field. PPL-Toolbox is publicly available on GitHub (https://github.com/versarchey/PPL-Toolbox).

Advances in three-dimensional (3D) genomics have revealed the spatial characteristics of chromatin interactions in gene expression regulation, which is crucial for understanding molecular mechanisms in biological processes. High-throughput technologies like ChIA-PET, Hi-C, and their derivatives methods have greatly enhanced our knowledge of 3D chromatin architecture. However, the chromatin interaction mechanisms remain largely unexplored. Deep learning, with its powerful feature extraction and pattern recognition capabilities, offers a promising approach for integrating multi-omics data, to build accurate predictive models of chromatin interaction matrices. This review systematically summarizes recent advances in chromatin interaction matrix prediction models. By integrating DNA sequences and epigenetic signals, we investigate the latest developments in these methods. This article details various models, focusing on how one-dimensional (1D) information transforms into the 3D structure chromatin interactions, and how the integration of different deep learning modules specifically affects model accuracy. Additionally, we discuss the critical role of DNA sequence information and epigenetic markers in shaping 3D genome interaction patterns. Finally, this review addresses the challenges in predicting chromatin interaction matrices, in order to improve the precise mapping of chromatin interaction matrices and DNA sequence, and supporting the transformation and theoretical development of 3D genomics across biological systems.

While convolutional neural network(CNN) have historically achieved significant success in 3D medical image segmentation, recent advancements in Vision Transformer (ViT) have led to increased popularity of Transformer architectures in this field. Leveraging the global information extraction capabilities of Transformers, we propose a novel parallel structure combining CNN and Transformer for multimodal brain tumor segmentation, named PTCNet. We introduce two key improvements to enhance multimodal segmentation:1) Introducing a novel channel attention module that replaces spatial self-attention in the Transformer to improve cross-modal information extraction.2) Enhancing the reversible feature extraction module based on residual structures to better detect small lesions. We evaluated our method on the BraTS 2023 brain metastasis dataset, achieving an average DSC of 61.42%, which is 1.12% higher than the second place. The innovative structure and improved modules of PTCNet significantly enhance performance in multimodal brain tumor segmentation tasks. Code and models are available at https://github.com/naiqizh/PTCNet.

In recent years, the Sustainable Development Goals (SDGs) have received increasing attention from scholars, an international agenda that aims to address global challenges such as poverty, hunger, and environmental damage. By accurately predicting soil moisture, we can effectively address the challenges of food security, sustainable management of water resources, conservation of ecosystems, and response to climate change. By improving soil moisture prediction techniques and methods, these global challenges can be solved more effectively. To better predict soil moisture at different levels, a new PSO-CNN-LSTM is proposed. This model combines the advantages of CNN-LSTM in extracting multivariate feature sequences and PSO in hyper-parametric global optimization. RNN, GRU, LSTM, CNN-GRU and CNN-LSTM are trained and tested as a control group on the publicly available dataset EARS-Land. The results demonstrated the better performance of PSO-CNN-LSTM in predicting soil moisture at different levels with different time output steps. Our study demonstrated that the proposed model could be a promising alternative for predicting soil moisture.

Single-cell multiomics techniques have been widely applied to detect the key signature of cells. These methods have achieved a single-molecule resolution and can even reveal spatial localization. These emerging methods provide insights elucidating the features of genomic, epigenomic and transcriptomic heterogeneity in individual cells. However, they have given rise to new computational challenges in data processing. Here, we describe Single-cell Single-molecule multiple Omics Pipeline (ScSmOP), a universal pipeline for barcode-indexed single-cell single-molecule multiomics data analysis. Essentially, the C language is utilized in ScSmOP to set up spaced-seed hash table-based algorithms for barcode identification according to ligation-based barcoding data and synthesis-based barcoding data, followed by data mapping and deconvolution. We demonstrate high reproducibility of data processing between ScSmOP and published pipelines in comprehensive analyses of single-cell omics data (scRNA-seq, scATAC-seq, scARC-seq), single-molecule chromatin interaction data (ChIA-Drop, SPRITE, RD-SPRITE), single-cell single-molecule chromatin interaction data (scSPRITE) and spatial transcriptomic data from various cell types and species. Additionally, ScSmOP shows more rapid performance and is a versatile, efficient, easy-to-use and robust pipeline for single-cell single-molecule multiomics data analysis.

The emerging ligation-free three-dimensional (3D) genome mapping technologies can identify multiplex chromatin interactions with single-molecule precision. These technologies not only offer new insight into high-dimensional chromatin organization and gene regulation, but also introduce new challenges in data visualization and analysis. To overcome these challenges, we developed MCIBox, a toolkit for multi-way chromatin interaction (MCI) analysis, including a visualization tool and a platform for identifying micro-domains with clustered single-molecule chromatin complexes. MCIBox is based on various clustering algorithms integrated with dimensionality reduction methods that can display multiplex chromatin interactions at single-molecule level, allowing users to explore chromatin extrusion patterns and super-enhancers regulation modes in transcription, and to identify single-molecule chromatin complexes that are clustered into micro-domains. Furthermore, MCIBox incorporates a two-dimensional kernel density estimation algorithm to identify micro-domains boundaries automatically. These micro-domains were stratified with distinctive signatures of transcription activity and contained different cell-cycle-associated genes. Taken together, MCIBox represents an invaluable tool for the study of multiple chromatin interactions and inaugurates a previously unappreciated view of 3D genome structure.

Topologically associated domains (TADs) are one of the important higher order chromatin structures with various sizes in the eukaryotic genomes. TAD boundaries, as the flanking regions between adjacent domains, can restrict the interactions of regulatory elements, including enhancers and promoters, and are generally dynamic and variable in different cells. However, the influence of sequence and epigenetic profile-based features in the identification of TAD boundaries is largely unknown. In this work, we proposed a method called pTADS (prediction of TAD boundary and strength), to predict TAD boundaries and boundary strength across multiple cell lines with DNA sequence and epigenetic profile information. The performance was assessed in seven cell lines and three TAD calling methods. The results demonstrate that the TAD boundary can be well predicted by the selected shared features across multiple cell lines. Especially, the model can be transferable to predict the TAD boundary from one cell line to other cell lines. The boundary strength can be characterized by boundary score with good performance. The predicted TAD boundary and TAD boundary strength are further confirmed by three Hi-C contact matrix-based methods across multiple cell lines. The codes and datasets are available at https://github.com/chrom3DEpi/pTADS.

MOTIVATION: Virus integration in the host genome is frequently reported to be closely associated with many human diseases, and the detection of virus integration is a critically challenging task. However, most existing tools show limited specificity and sensitivity. Therefore, the objective of this study is to develop a method for accurate detection of virus integration into host genomes. RESULTS: Herein, we report a novel method termed HIVID2 that is a significant upgrade of HIVID. HIVID2 performs a paired-end combination (PE-combination) for potentially integrated reads. The resulting sequences are then remapped onto the reference genomes, and both split and discordant chimeric reads are used to identify accurate integration breakpoints with high confidence. HIVID2 represents a great improvement in specificity and sensitivity, and predicts breakpoints closer to the real integrations, compared with existing methods. The advantage of our method was demonstrated using both simulated and real datasets. HIVID2 uncovered novel integration breakpoints in well-known cervical cancer-related genes, including FHIT and LRP1B, which was verified using protein expression data. In addition, HIVID2 allows the user to decide whether to automatically perform advanced analysis using the identified virus integrations. By analyzing the simulated data and real data tests, we demonstrated that HIVID2 is not only more accurate than HIVID but also better than other existing programs with respect to both sensitivity and specificity. We believe that HIVID2 will help in enhancing future research associated with virus integration. AVAILABILITYAND IMPLEMENTATION: HIVID2 can be accessed at https://github.com/zengxi-hada/HIVID2/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The eight papers in this special section were presented at the 17th Asia Pacific Bioinformatics Conference (APBC), which was held in Wuhan, China, 14-16 January 2019.

BACKGROUND: DNA methylation is an important epigenetic modification that plays a critical role in most eukaryotic organisms. Parental alleles in haploid genomes may exhibit different methylation patterns, which can lead to different phenotypes and even different therapeutic and drug responses to diseases. However, to our knowledge, no software is available for the identification of DNA methylation haplotype regions with combined allele-specific DNA methylation, single nucleotide polymorphisms (SNPs) and high-throughput chromosome conformation capture (Hi-C) data. RESULTS: In this paper, we developed a new method, MethHaplo, that identify DNA methylation haplotype regions with allele-specific DNA methylation and SNPs from whole-genome bisulfite sequencing (WGBS) data. Our results showed that methylation haplotype regions were ten times longer than haplotypes with SNPs only. When we integrate WGBS and Hi-C data, MethHaplo could call even longer haplotypes. CONCLUSIONS: This study illustrates the usefulness of methylation haplotypes. By constructing methylation haplotypes for various cell lines, we provide a clearer picture of the effect of DNA methylation on gene expression, histone modification and three-dimensional chromosome structure at the haplotype level. Our method could benefit the study of parental inheritance-related disease and hybrid vigor in agriculture.

BACKGROUND: DNA methylation plays crucial roles in most eukaryotic organisms. Bisulfite sequencing (BS-Seq) is a sequencing approach that provides quantitative cytosine methylation levels in genome-wide scope and single-base resolution. However, genomic variations such as insertions and deletions (indels) affect methylation calling, and the alignment of reads near/across indels becomes inaccurate in the presence of polymorphisms. Hence, the simultaneous detection of DNA methylation and indels is important for exploring the mechanisms of functional regulation in organisms. RESULTS: These problems motivated us to develop the algorithm BatMeth2, which can align BS reads with high accuracy while allowing for variable-length indels with respect to the reference genome. The results from simulated and real bisulfite DNA methylation data demonstrated that our proposed method increases alignment accuracy. Additionally, BatMeth2 can calculate the methylation levels of individual loci, genomic regions or functional regions such as genes/transposable elements. Additional programs were also developed to provide methylation data annotation, visualization, and differentially methylated cytosine/region (DMC/DMR) detection. The whole package provides new tools and will benefit bisulfite data analysis. CONCLUSION: BatMeth2 improves DNA methylation calling, particularly for regions close to indels. It is an autorun package and easy to use. In addition, a DNA methylation visualization program and a differential analysis program are provided in BatMeth2. We believe that BatMeth2 will facilitate the study of the mechanisms of DNA methylation in development and disease. BatMeth2 is an open source software program and is available on GitHub ( https://github.com/GuoliangLi-HZAU/BatMeth2 /).

In this paper, we exploit deep convolutional features for object appearance modeling and propose a simple while effective deep discriminative model (DDM) for visual tracking. The proposed DDM takes as input the deep features and outputs an object-background confidence map. Considering that both spatial information from lower convolutional layers and semantic information from higher layers benefit object tracking, we construct multiple deep discriminative models (DDMs) for each layer and combine these confidence maps from each layer to obtain the final object-background confidence map. To reduce the risk of model drift, we propose to adopt a saliency method to generate object candidates. Object tracking is then achieved by finding the candidate with the largest confidence value. Experiments on a large-scale tracking benchmark demonstrate that the propose method performs favorably against state-of-the-art trackers.

Translation initiation sites (TISs) are important signals in cDNA sequences. Many research efforts have tried to predict TISs in cDNA sequences. In this paper, we propose to use mixture Gaussian models for TIS prediction. Using both local features and some features generated from global measures, the proposed method predicts TISs with a sensitivity of 98 percent and a specificity of 93.6 percent. Our method outperforms many other existing methods in sensitivity while keeping specificity high. We attribute the improvement in sensitivity to the nature of the global features and the mixture Gaussian models.