Applying interpretable machine learning in computational biology-pitfalls, recommendations and opportunities for new developments.

Recent advances in machine learning have enabled the development of next-generation predictive models for complex computational biology problems, thereby spurring the use of interpretable machine learning (IML) to unveil biological insights. However, guidelines for using IML in computational biology are generally underdeveloped. We provide an overview of IML methods and evaluation techniques and discuss common pitfalls encountered when applying IML methods to computational biology problems. We also highlight open questions, especially in the era of large language models, and call for collaboration between IML and computational biology researchers.

Computational methods for analysing multiscale 3D genome organization.

Recent progress in whole-genome mapping and imaging technologies has enabled the characterization of the spatial organization and folding of the genome in the nucleus. In parallel, advanced computational methods have been developed to leverage these mapping data to reveal multiscale three-dimensional (3D) genome features and to provide a more complete view of genome structure and its connections to genome functions such as transcription. Here, we discuss how recently developed computational tools, including machine-learning-based methods and integrative structure-modelling frameworks, have led to a systematic, multiscale delineation of the connections among different scales of 3D genome organization, genomic and epigenomic features, functional nuclear components and genome function. However, approaches that more comprehensively integrate a wide variety of genomic and imaging datasets are still needed to uncover the functional role of 3D genome structure in defining cellular phenotypes in health and disease.

Interkingdom Hormonal Regulations between Plants and Animals Provide New Insight into Food Safety.

Food safety has become an attractive topic among consumers. Raw material production for food is also a focus of social attention. As hormones are widely used in agriculture and human disease control, consumers' concerns about the safety of hormone agents have never disappeared. The present review focuses on the interkingdom regulations of exogenous animal hormones in plants and phytohormones in animals, including physiology and stress resistance. We summarize these interactions to give the public, researchers, and policymakers some guidance and suggestions. Accumulated evidence demonstrates comprehensive hormonal regulation across plants and animals. Animal hormones, interacting with phytohormones, help regulate plant development and enhance environmental resistance. Correspondingly, phytohormones may also cause damage to the reproductive and urinary systems of animals. Notably, the disease-resistant role of phytohormones is revealed against neurodegenerative diseases, cardiovascular disease, cancer, and diabetes. These resistances derive from the control for abnormal cell cycle, energy balance, and activity of enzymes. Further exploration of these cross-kingdom mechanisms would surely be of greater benefit to human health and agriculture development.

UNADON: transformer-based model to predict genome-wide chromosome spatial position.

MOTIVATION: The spatial positioning of chromosomes relative to functional nuclear bodies is intertwined with genome functions such as transcription. However, the sequence patterns and epigenomic features that collectively influence chromatin spatial positioning in a genome-wide manner are not well understood. RESULTS: Here, we develop a new transformer-based deep learning model called UNADON, which predicts the genome-wide cytological distance to a specific type of nuclear body, as measured by TSA-seq, using both sequence features and epigenomic signals. Evaluations of UNADON in four cell lines (K562, H1, HFFc6, HCT116) show high accuracy in predicting chromatin spatial positioning to nuclear bodies when trained on a single cell line. UNADON also performed well in an unseen cell type. Importantly, we reveal potential sequence and epigenomic factors that affect large-scale chromatin compartmentalization in nuclear bodies. Together, UNADON provides new insights into the principles between sequence features and large-scale chromatin spatial localization, which has important implications for understanding nuclear structure and function. AVAILABILITY AND IMPLEMENTATION: The source code of UNADON can be found at https://github.com/ma-compbio/UNADON.

UNADON: Transformer-based model to predict genome-wide chromosome spatial position.

The spatial positioning of chromosomes relative to functional nuclear bodies is intertwined with genome functions such as transcription. However, the sequence patterns and epigenomic features that collectively influence chromatin spatial positioning in a genome-wide manner are not well understood. Here, we develop a new transformer-based deep learning model called UNADON, which predicts the genome-wide cytological distance to a specific type of nuclear body, as measured by TSA-seq, using both sequence features and epigenomic signals. Evaluations of UNADON in four cell lines (K562, H1, HFFc6, HCT116) show high accuracy in predicting chromatin spatial positioning to nuclear bodies when trained on a single cell line. UNADON also performed well in an unseen cell type. Importantly, we reveal potential sequence and epigenomic factors that affect large-scale chromatin compartmentalization to nuclear bodies. Together, UNADON provides new insights into the principles between sequence features and large-scale chromatin spatial localization, which has important implications for understanding nuclear structure and function.

Changes in cropland soil carbon through improved management practices in China: A meta-analysis.

Recommended management practices (RMPs, e.g., manuring, no-tillage, crop residue return) can increase soil organic carbon (SOC), reduce greenhouse gas emissions, and maintain soil health in croplands. However, there is no consensus on how RMPs affect the SOC storage potential of cropland soils for climate change mitigation. Here, based on 2301 comparisons from 158 peer-reviewed papers, a meta-analysis was conducted to explore management-induced SOC stock changes and their variations under different conditions. The results show that SOC stocks in the 0-20 cm layer were increased by 31.8% when chemical fertilization combined with manure application was compared with no fertilizer; 9.98% when no-tillage was compared with plow tillage; and 10.84% when straw return was compared with removal. The RMPs favorably increased SOC stock in arid areas, and in alkaline and fine-textured soils. Initial SOC, carbon-nitrogen ratio, and experimental duration could also affect SOC storage. Compared with the initial SOC stock, RMPs increased the SOC sequestration potential by 2.6-4.5% in the 0-20 cm soil depth, indicating that these practices can help China achieve targets to increase SOC by 4.0‰. Hence, it is essential to implement RMPs for climate change mitigation and soil fertility improvement.

Effects of tillage and cropping sequences on crop production and environmental benefits in the North China Plain.

The ever-increasing trend of greenhouse gas (GHG) emissions is accelerating global warming and threatening food security. Environmental benefits and sustainable food production must be pursued locally and globally. Thus, a field experiment was conducted in 2015 to understand how to balance the trade-offs between agronomic productivity and environment quality in the North China Plain (NCP). Eight treatments consisted of two factors, i.e., (1) tillage practices: rotary tillage (RT) and no-till (NT), and (2) cropping sequences (CS): maize-wheat-soybean-wheat (MWSW), soybean-wheat-maize-wheat (SWMW), soybean-wheat (SW), and maize-wheat (MW). The economic and environmental benefits were evaluated by multiple indicators including the carbon footprint (CF), maize equivalent economic yield (MEEY), energy yield (EY), and carbon sustainability index (CSI). Compared with NT, RT increased the EY and MEEY, but emitted 9.4% higher GHGs. Among different CSs, no significant reduction was observed in CF. The lowest (2.0 Mg CO2-eq ha-1 year-1) and the highest (5.6 Mg CO2-eq ha-1 year-1) CF values were observed under MW and SWMW, respectively. However, CSs with soybean enhanced MEEY and the net revenue due to their higher price compared to that of MW. Although the highest CSI was observed under RT-MW, soybean-based crop rotation could offset the decline in CSI under NT when compared to that for RT. These findings suggest that conservation agriculture (CA) could enhance the balance in trade-offs between economic and environmental benefits. Additional research is needed on how to achieve high crop production by establishing a highly efficient CA system in the NCP.

Machine Learning Methods for Exploring Sequence Determinants of 3D Genome Organization.

In higher eukaryotic cells, chromosomes are folded inside the nucleus. Recent advances in whole-genome mapping technologies have revealed the multiscale features of 3D genome organization that are intertwined with fundamental genome functions. However, DNA sequence determinants that modulate the formation of 3D genome organization remain poorly characterized. In the past few years, predicting 3D genome organization based on DNA sequence features has become an active area of research. Here, we review the recent progress in computational approaches to unraveling important sequence elements for 3D genome organization. In particular, we discuss the rapid development of machine learning-based methods that facilitate the connections between DNA sequence features and 3D genome architectures at different scales. While much progress has been made in developing predictive models for revealing important sequence features for 3D genome organization, new research is urgently needed to incorporate multi-omic data and enhance model interpretability, further advancing our understanding of gene regulation mechanisms through the lens of 3D genome organization.

[Effects of PD-L1 on immunosuppression of bacterial sepsis and its relevant mechanism].

Objective: To investigate the expression of programmed death ligand-1 (PD-L1) in dendritic cells (DCS) and its related signaling pathway in lipopolysaccharide (LPS)-induced immunosuppression of bacterial sepsis.Methods: Stimulating with bacterial LPS, bone marrow-derived dendritic cells could induce T lymphocyte immunosuppression imitating bacterial sepsis model. The experiments were divided into 5 groups: control group, LPS group, 2-(4-morpholinyl)-8-phenyl-4H-1- benzopyran-4-one (LY294002)+LPS group, pyrrolidinedithiocarbamate(PDTC)+LPS group and LPS+anti-PD-L1 group with 6 multiple wells in each group. After mice bone marrow source monocytes were cultured with rmGM-CSF (10 ng/ml) and rmIL-4 (1 ng/ml) in 10% fetal bovine serum 1640 for 4 days, DCs cells were treated with with 10 ng/ml LPS for 12 h to obtain immunosuppressive cells with high expression of PD-L1. Pathway-inhibitors LY294002 (10 µmol/L) and PDTC (20 µmol/L) were used to block PI3K and NF-κB signals. Flow cytometry and confocal laser scanning microscopy were used to detect the PD-L1 expression and phosphatidylinositol 3 kinase/protein kinase B (PI3K/AKT) signal activation on DCs. BrdU cell proliferation assay and γ-interferon enzyme-linked immunospot assay were used to detect ovalbumin specific T lymphocyte proliferation response and cytotoxic T cell response, respectively. Results: Compared with the control group, the percentage of PD-L1 positive cells and PD-L1 red fluorescence intensity of DCs were all increased(P＜0.01), while DCs- mediated T cell proliferation and γ-interferon spot-forming cell number were decreased (P＜0.01).PI3K inhibitor LY294002, NF-κB inhibitor PDTC and PD-L1 blocking antibody could significantly reverse the inhibition of DCs mediated T lymphocytes immunosuppression above (P＜0.01). Conclusion: PD-L1 was a key molecule that mediates immunosuppression in lipopolysaccharide induced bacterial sepsis. PI3K Signal and NF- κB signal were also involved in this immunosuppressive process.

Cancer mutational signatures representation by large-scale context embedding.

MOTIVATION: The accumulation of somatic mutations plays critical roles in cancer development and progression. However, the global patterns of somatic mutations, especially non-coding mutations, and their roles in defining molecular subtypes of cancer have not been well characterized due to the computational challenges in analysing the complex mutational patterns. RESULTS: Here, we develop a new algorithm, called MutSpace, to effectively extract patient-specific mutational features using an embedding framework for larger sequence context. Our method is motivated by the observation that the mutation rate at megabase scale and the local mutational patterns jointly contribute to distinguishing cancer subtypes, both of which can be simultaneously captured by MutSpace. Simulation evaluations show that MutSpace can effectively characterize mutational features from known patient subgroups and achieve superior performance compared with previous methods. As a proof-of-principle, we apply MutSpace to 560 breast cancer patient samples and demonstrate that our method achieves high accuracy in subtype identification. In addition, the learned embeddings from MutSpace reflect intrinsic patterns of breast cancer subtypes and other features of genome structure and function. MutSpace is a promising new framework to better understand cancer heterogeneity based on somatic mutations. AVAILABILITY AND IMPLEMENTATION: Source code of MutSpace can be accessed at: https://github.com/ma-compbio/MutSpace. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.