Bromine Substitution Improves the Photothermal Performance of π-Conjugated Phototheranostic Molecules.

NIR-II fluorescence imaging-guided photothermal therapy (PTT) has been widely investigated due to its great application potential in tumor theranostics. PTT is an effective and non-invasive tumor treatment method that can adapt to tumor hypoxia; nevertheless, simple and effective strategies are still desired to develop new materials with excellent PTT properties to meet clinical requirements. In this work, we developed a bromine-substitution strategy to enhance the PTT of A-D-A'-D-A π-conjugated molecules. The experimental results reveal that bromine substitution can notably enhance the absorptivity (ϵ) and photothermal conversion efficiency (PCE) of the π-conjugated molecules, resulting in the brominated molecules generating two times more heat (ϵ808 nm ×PCE) than their unsubstituted counterpart. We disclose that the enhanced photothermal properties of bromine-substituted π-conjugated molecules are a combined outcome of the heavy-atom effect, enhanced ICT effect, and more intense bromine-mediate intermolecular π-π stacking. Finally, the NIR-II tumor imaging capability and efficient PTT tumor ablation of the brominated π-conjugated materials demonstrate that bromine substitution is a promising strategy for developing future high-performance NIR-II imaging-guided PTT agents.

Polymer-DNA Carriers Co-Deliver Photosensitizer and siRNA for Light-Promoted Gene Transfection and Hypoxia-Relieved Photodynamic Therapy.

Photochemical internalization is an efficient strategy relying on photodynamic reactions to promote siRNA endosomal escape for the success of RNA-interference gene regulation, which makes gene-photodynamic combined therapy highly synergistic and efficient. However, it is still desired to explore capable carriers to improve the delivery efficiency of the immiscible siRNA and organic photosensitizers simultaneously. Herein, we employ a micellar nanostructure (PSNA) self-assembled from polymer-DNA molecular chimeras to fulfill this task. PSNA can plentifully load photosensitizers in its hydrophobic core simply by the nanoprecipitation method. Moreover, it can organize siRNA self-assembly by the densely packed DNA shell, which leads to a higher loading capacity than the typical electrostatic condensation method. The experimental results prove that this PSNA carrier can greatly facilitate siRNA escape from the endosome/lysosome and enhance transfection. Accordingly, the PSNA-administrated therapy exhibits a significantly improved anti-tumor efficacy owing to the highly efficient co-delivery capability.

CAFU: a Galaxy framework for exploring unmapped RNA-Seq data.

A widely used approach in transcriptome analysis is the alignment of short reads to a reference genome. However, owing to the deficiencies of specially designed analytical systems, short reads unmapped to the genome sequence are usually ignored, resulting in the loss of significant biological information and insights. To fill this gap, we present Comprehensive Assembly and Functional annotation of Unmapped RNA-Seq data (CAFU), a Galaxy-based framework that can facilitate the large-scale analysis of unmapped RNA sequencing (RNA-Seq) reads from single- and mixed-species samples. By taking advantage of machine learning techniques, CAFU addresses the issue of accurately identifying the species origin of transcripts assembled using unmapped reads from mixed-species samples. CAFU also represents an innovation in that it provides a comprehensive collection of functions required for transcript confidence evaluation, coding potential calculation, sequence and expression characterization and function annotation. These functions and their dependencies have been integrated into a Galaxy framework that provides access to CAFU via a user-friendly interface, dramatically simplifying complex exploration tasks involving unmapped RNA-Seq reads. CAFU has been validated with RNA-Seq data sets from wheat and Zea mays (maize) samples. CAFU is freely available via GitHub: https://github.com/cma2015/CAFU.

Inhibitory effect of ficin on Candida albicans biofilm formation and pre-formed biofilms.

BACKGROUND: To investigate the effect of ficin, a type of proteases, on Candida albicans (C. albicans) biofilm, including forming and pre-formed biofilms. METHODS: Crystal violet tests together with colony forming unit (CFU) counts were used to detect fungal biofilm biomass. Live/dead staining of biofilms observed by confocal laser scanning microscopy was used to monitor fungal activity. Finally, gene expression of C. albicans within biofilms was assessed by qRT-PCR. RESULTS: According to our results, biofilm biomass was dramatically reduced by ficin in both biofilm formation and pre-formed biofilms, as revealed by the crystal violet assay and CFU count (p < 0.05). Fungal activity in biofilm formation and pre-formed biofilms was not significantly influenced by ficin according to live/dead staining. Fungal polymorphism and biofilm associated gene expression were influenced by ficin, especially in groups with prominent antibiofilm effects. CONCLUSIONS: In summary, ficin effectively inhibited C. albicans biofilm formation and detached its preformed biofilm, and it might be used to treat C. albicans biofilm associated problems.

CellSim: a novel software to calculate cell similarity and identify their co-regulation networks.

BACKGROUND: Cell direct reprogramming technology has been rapidly developed with its low risk of tumor risk and avoidance of ethical issues caused by stem cells, but it is still limited to specific cell types. Direct reprogramming from an original cell to target cell type needs the cell similarity and cell specific regulatory network. The position and function of cells in vivo, can provide some hints about the cell similarity. However, it still needs further clarification based on molecular level studies. RESULT: CellSim is therefore developed to offer a solution for cell similarity calculation and a tool of bioinformatics for researchers. CellSim is a novel tool for the similarity calculation of different cells based on cell ontology and molecular networks in over 2000 different human cell types and presents sharing regulation networks of part cells. CellSim can also calculate cell types by entering a list of genes, including more than 250 human normal tissue specific cell types and 130 cancer cell types. The results are shown in both tables and spider charts which can be preserved easily and freely. CONCLUSION: CellSim aims to provide a computational strategy for cell similarity and the identification of distinct cell types. Stable CellSim releases (Windows, Linux, and Mac OS/X) are available at: www.cellsim.nwsuaflmz.com , and source code is available at: https://github.com/lileijie1992/CellSim/ .

Well-Optimized Conjugated GO-DNA Nanosystem for Sensitive Ratiometric pH Detection in Live Cells.

Intracellular pH is a vital parameter which can reflect the physiological process, and the detection of intracellular pH with a high signal-to-noise ratio (SNR) remains a challenge. Compared to pH biosensors based on a single-wavelength signal, it is much easier to obtain better sensitivity and higher SNR from the biosensors by two-wavelength ratiometric signals. In this study, we used DNA-grafted graphene oxide (GO) to ratiometrically detect intracellular pH ranging from basic to acidic. A high SNR with a 35-fold difference in the ratiometric output has been achieved through careful optimization: (1) A high DNA conjugation yield of 45% has been gained through utilizing the partial double-stranded assembly strategy. (2) Herring sperm DNA (HSD) plays an important role in improving the sensitivity of the nanosystem by purifying and passivating the surface of GO; therefore, the concentration of HSD has been optimized to pursue the most sensitive ratiometric response. Apart from the ultrahigh SNR, fabricated GO-AR-Cy5/IFO-Cy3 exhibited excellent stability and biocompatibility in biological environments. Further experiments demonstrated that the nanosystem worked well in live cells in response to pH changes. It is possible to distinguish small pH differences and realize quantitative detection based on ratiometric fluorescence imaging by laser scanning confocal microscope analysis, which makes the nanosystem a promising candidate for further biological study and clinical applications.

An atypical R2R3 MYB transcription factor increases cold hardiness by CBF-dependent and CBF-independent pathways in apple.

Apple (Malus × domestica) trees are vulnerable to freezing temperatures. However, there has been only limited success in developing cold-hardy cultivars. This lack of progress is due at least partly to lack of understanding of the molecular mechanisms of freezing tolerance in apple. In this study, we evaluated the potential roles for two R2R3 MYB transcription factors (TFs), MYB88 and the paralogous FLP (MYB124), in cold stress in apple and Arabidopsis. We found that MYB88 and MYB124 positively regulate freezing tolerance and cold-responsive gene expression in both apple and Arabidopsis. Chromatin-Immunoprecipitation-qPCR and electrophoretic mobility shift assays showed that MdMYB88/MdMYB124 act as direct regulators of the COLD SHOCK DOMAIN PROTEIN 3 (MdCSP3) and CIRCADIAN CLOCK ASSOCIATED 1 (MdCCA1) genes. Dual luciferase reporter assay indicated that MdCCA1 but not MdCSP3 activated the expression of MdCBF3 under cold stress. Moreover, MdMYB88 and MdMYB124 promoted anthocyanin accumulation and H2 O2 detoxification in response to cold. Taken together, our results suggest that MdMYB88 and MdMYB124 positively regulate cold hardiness and cold-responsive gene expression under cold stress by C-REPEAT BINDING FACTOR (CBF)-dependent and CBF-independent pathways.

pH-Responsive Graphene Oxide-DNA Nanosystem for Live Cell Imaging and Detection.

The interaction between graphene oxide (GO) and DNA is very sensitive to the environment. For example, under acidic conditions, the affinity of GO for DNA is enhanced, weakening the capability of GO to distinguish DNAs with different conformations. This effect has impeded the development of sensitive pH biosensors based on GO-DNA nanosystems. In this work, we systematically studied the affinity between GO and i-motif forming oligonucleotides (IFOs) at different pH values and developed a herring sperm DNA (HSD) treatment method. Using this method, HSD occupies the surface of GO, compromising the attractive force of GO that is significantly enhanced under acidic conditions. As a result, the ability of GO to distinguish between "open" and "closed" IFOs is successfully generalized to a wider pH range. Finally, a pH-sensitive GO-IFO nanosystem was fabricated that showed excellent sensing ability both in vitro and for intracellular pH detection. Because the interaction between GO and DNA is the basis for constructing GO-DNA biosensors, the strategy developed in this work shows great potential to be applied in a variety of other GO-DNA sensing systems.

MutS HOMOLOG1-derived epigenetic breeding potential in tomato.

Evidence is compelling in support of a naturally occurring epigenetic influence on phenotype expression in land plants, although discerning the epigenetic contribution is difficult. Agriculturally important attributes like heterosis, inbreeding depression, phenotypic plasticity, and environmental stress response are thought to have significant epigenetic components, but unequivocal demonstration of this is often infeasible. Here, we investigate gene silencing of a single nuclear gene, MutS HOMOLOG1 (MSH1), in the tomato (Solanum lycopersicum) 'Rutgers' to effect developmental reprogramming of the plant. The condition is heritable in subsequent generations independent of the MSH1-RNA interference transgene. Crossing these transgene-null, developmentally altered plants to the isogenic cv Rutgers wild type results in progeny lines that show enhanced, heritable growth vigor under both greenhouse and field conditions. This boosted vigor appears to be graft transmissible and is partially reversed by treatment with the methylation inhibitor 5-azacytidine, implying the influence of mobile, epigenetic factors and DNA methylation changes. These data provide compelling evidence for the feasibility of epigenetic breeding in a crop plant.

Biogeographic and co-occurrence network differentiation of fungal communities in warm-temperate montane soils.

Studying the biogeographic patterns of fungal communities across altitudinal and soil depth gradients is essential for understanding how environmental variations shape the diversity and functionality of these complex ecological assemblages. Here, we evaluated the response and assembly patterns of fungal communities to altitude and soil depth, and the co-occurrence patterns influencing soil fungal metabolic preferences on Dongling Mountain. We observed significant variations in fungal ß-diversity, driven by elevation and soil depth, with climatic parameters (MAT and MAP) and nutrient concentrations (TOC, TP, and TN) serving as prominent influencers. Additionally, we found that the multiple substrate-induced respiration rate of fungi degrading various carbon substrates was diminished in high-altitude and subsurface soils compared to low-altitude and surface soils. Stochastic processes play a more important role in controlling fungal community assembly than deterministic processes, with dispersal limitation emerging as the main driver of community assembly. While greater network complexity was evident in the topsoil compared to the subsoil, both layers harbored altitude-sensitive OTUs (asOTUs) that belonging to distinct modules. Moreover, fungal groups sensitive to the same altitude exhibited similar metabolic preferences. The asOTUs designated for lower altitude areas favored unstable carbon substrates (glucose and sucrose), while those designated as higher altitude areas exhibited a preference for recalcitrant carbon (xylan and lignin). This evidence suggests that soil fungal communities respond to environmental changes by trading off their life strategies and metabolic characteristics.

The first emergence of paramyxovirus type 12 in wild birds in mainland, China.

Avian paramyxoviruses (APMV) belong to the subfamily Avulavirinae of the family Paramyxoviridae and include 22 distinct subtypes or serotypes (1-22). Avian paramyxovirus serotype 12 (APMV-12) is found sporadically in wild birds worldwide, and reports from only Italy and Taiwan have been published to date; information on its genetic variation and biological characteristics is still limited. In this study, 3 APMV-12 strains, designated WB19, LY9, and LY11, were isolated from 8643 wild bird faecal samples during the annual influenza virus surveillance of wild birds in Guangdong, China between 2018 and 2024, which is first reported in mainland China. The complete genomes of the 3 viruses with 6 gene segments, 3'-N-P-M-F-HN-L-5', were 15,231 nt in length. Phylogenetic analysis based on the whole genome showed that the 3 APMV-12 strains had the highest homology with an APMV-12 strain isolated from Taiwan in 2015, followed by the prototype APMV-12 strains isolated from mallard ducks in Italy in 2005. Genetic analysis of the whole gene of each of them indicated that they were derived from a Eurasian lineage. This study provides additional evidence that wild birds transmit viruses between countries, and this should be monitored to understand APMV transmission, evolution and epidemiology.

Biologically produced and metal-organic framework delivered dual-cut CRISPR/Cas9 system for efficient gene editing and sensitized cancer therapy.

Manipulation of the lactate metabolism is an efficient way for cancer treatment given its involvement in cancer development, metastasis, and immune escape. However, most of the inhibitors of lactate transport carriers suffer from poor specificity. Herein, we use the CRISPR/Cas9 system to precisely downregulate the monocarboxylate carrier 1 (MCT1) expression. To avoid the self-repairing during the gene editing process, a dual-Cas9 ribonucleoproteins (duRNPs) system is generated using the biological fermentation method and delivered into cells by the zeolitic imidazolate framework-8 (ZIF-8) nanoparticles, enabling precise removal of a specific DNA fragment from the genome. For efficient cancer therapy, a specific glucose transporter 1 inhibitor (BAY-876) is co-delivered with the duRNPs, forming BAY/duRNPs@ZIF-8 nanoparticle. ZIF-8 nanoparticles can deliver the duRNPs into cells within 1 h, which efficiently downregulates the MCT1 expression, and prohibits lactate influx. Through simultaneous inhibition of the lactate and glucose influx, BAY/duRNPs@ZIF-8 prohibits ATP generation, arrests cell cycle, inhibits cell proliferation, and finally induces cellular apoptosis both in vitro and in vivo. Consequently, we demonstrate that the biologically produced duRNPs delivered into cells by the nonviral ZIF-8 carrier have expanded the CRISPR/Cas gene editing toolbox and elevated the gene editing efficiency, which will promote biological studies and clinical applications. STATEMENT OF SIGNIFICANCE: The CRISPR/Cas9 system, widely used as an efficient gene editing tool, faces a challenge due to cells' ability to self-repair. To address this issue, a strategy involving dual-cutting of the genome DNA has been designed and implemented. This strategy utilizes biologically produced dual-ribonucleoproteins delivered by a metal-organic framework. The effectiveness of this dual-cut CRISPR-Cas9 system has been demonstrated through a therapeutic approach targeting the simultaneous inhibition of lactate and glucose influx in cancer cells. The utilization of the dual-cut gene editing strategy has provided valuable insights into gene editing and expanded the toolbox of the CRISPR/Cas-based gene editing system. It has the potential to enable more efficient and precise manipulation of specific protein expression in the future.

Depth effects on bacterial community altitudinal patterns and assembly processes in the warm-temperate montane forests of China.

Soil bacterial communities are essential for ecosystem function, yet their response along altitudinal gradients in different soil strata remains unclear. Understanding bacterial community co-occurrence networks and assembly patterns in mountain ecosystems is crucial for comprehending microbial ecosystem functions. We utilized Illumina MiSeq sequencing to study bacterial diversity and assembly patterns of surface and subsurface soils across a range of elevations (700 to 2100 m) on Dongling Mountain. Our results showed significant altitudinal distribution patterns concerning bacterial diversity and structure in the surface soil. The bacterial diversity exhibited a consistent decrease, while specific taxa demonstrated unique patterns along the altitudinal gradient. However, no altitudinal dependence was observed for bacterial diversity and community structure in the subsurface soil. Additionally, a shift in bacterial ecological groups is evident with changing soil depth. Copiotrophic taxa thrive in surface soils characterized by higher carbon and nutrient content, while oligotrophic taxa dominate in subsurface soils with more limited resources. Bacterial community characteristics exhibited strong correlations with soil organic carbon in both soil layers, followed by pH in the surface soil and soil moisture in the subsurface soil. With increasing depth, there is an observable increase in taxa-taxa interaction complexity and network structure within bacterial communities. The surface soil exhibits greater sensitivity to environmental perturbations, leading to increased modularity and an abundance of positive relationships in its community networks compared to the subsurface soil. Furthermore, the bacterial community at different depths was influenced by combining deterministic and stochastic processes, with stochasticity (homogenizing dispersal and undominated) decreasing and determinism (heterogeneous selection) increasing with soil depth.

Ultra-stable MOF@MOF nanoplatform for photodynamic therapy sensitized by relieved hypoxia due to mitochondrial respiration inhibition.

Metal-organic frameworks (MOFs) with periodically arranged porphyrinic linkers avoiding the self-quenching issue of porphyrins in photodynamic therapy (PDT) have been widely applied. However, the porphyrinic MOFs still face challenges of poor stability under physiological conditions and limited photodynamic efficiency by the hypoxia condition of tumors. Herein, we fabricate the MOF@MOF structure with a protective MOF shell to improve the stability and relieve the hypoxia condition of tumors for sensitized PDT. Under protection of the MOF shell, the MOF@MOF structure can keep intact for 96 h under physiological conditions. Consequently, the tumoral accumulation efficiency is two folds of the MOF core. Furthermore, the MOF shell decomposes under acidic environment, and the loaded inhibitor of mitochondria pyruvate carrier (7-amino carboxycoumarins-2, 7ACC2) will be released. 7ACC2 inhibits the mitochondrial pyruvate influx and simultaneously blocks glucose and lactate from fueling the mitochondrial respiration, thereupon relieving the hypoxia condition of tumors. Under a 5-min laser irradiation, the 7ACC2 carrying MOF@MOF nanoplatforms induced doubled cellular apoptosis and reduced 70% of the tumor growth compared with the cargo-free MOF@MOF. In summary, the design of this stable and hypoxia self-relievable MOF@MOF nanoplatform will enlighten the future development of MOF-based nanomedicines and PDT. STATEMENT OF SIGNIFICANCE: Though widely used for photodynamic therapy (PDT) in previous studies, porphyrinic metal-organic frameworks (MOFs) still face challenges in poor stability under physiological conditions and limited photodynamic efficiency due to the hypoxia condition of tumors. In order to solve these problems, (1) we develop the MOF@MOF strategy to improve the physiological stability; (2) an inhibitor of mitochondria pyruvate carrier, 7-amino carboxycoumarins-2 (7ACC2), is loaded to inhibit the mitochondrial pyruvate influx and simultaneously block glucose and lactate from fueling the mitochondrial respiration, thereupon relieving the hypoxia condition of tumors. In comparison with previous studies, our strategy simultaneously improves stability and overcomes the limited PDT efficiency in the hypoxia tumor tissue, which will enlighten the future development of MOF-based nanomedicines and PDT.

Concern regarding H3-subtype avian influenza virus.

The H3-subtype of avian influenza virus (AIV) is one of the most frequently detected low pathogenic avian influenza virus (LPAIV) subtypes in birds and fowls, causing substantial economic loss to the poultry industry. Most importantly, besides poultry, mammals could also be infected with it, such as swines, canines, equines, felines, and humans, posing a serious public health threat. This allows the virus to persist widely in poultry and wild birds for a long time, where it may mix with other subtypes, providing conditions for viral recombination or reassortment. Currently, the monitoring of H3-subtype AIV is inadequate, and there is a lack of effective prevention and control measures for H3-subtype AIV. Here, the epidemiology, phylogeny, and genetic variation of H3-subtype AIV were analyzed, and nonsynonymous and synonymous substitution rates (dN/dS) were calculated. Through these steps, we aimed to clarify the current epidemiological feature and evolutionary characteristics of H3-subtype AIV, and provide an operative reference for future scientific control of H3-subtype AIV.

NIR-II Luminescent and Multi-Responsive Rare Earth Nanocrystals for Improved Chemodynamic Therapy.

Chemodynamic therapy (CDT) based on the Fe2+-mediated Fenton reaction can amplify intracellular oxidative stress by producing toxic •OH. However, the high-dose need for Fe2+ delivery in tumors and its significant cytotoxicity to normal tissues set a challenge. Therefore, a controllable delivery to activate the Fenton reaction and enhance Fe2+ tumor accumulation has become an approach to solve this conflict. Herein, we report a rare-earth-nanocrystal (RENC)-based Fe2+ delivery system using light-control techniques and DNA nanotechnology to realize programmable Fe2+ delivery. Ferrocenes, the source of Fe2+, are modified on the surface of RENCs through pH-responsive DNAs, which are further shielded by a PEG layer to elongate blood circulation and "turn off" the cytotoxicity of ferrocene. The up-/down-conversion dual-mode emissions of RENCs endow the delivery system with both capabilities of diagnosis and delivery control. The down-conversion NIR-II fluorescence can locate tumors. Consequently, up-conversion UV light spatiotemporally activates the catalytic activity of Fe2+ by shedding off the protective PEG layer. The exposed ferrocene-DNAs not only can "turn on" Fenton catalytic activity but also respond to tumor acidity, driving cross-linking and enhanced Fe2+ enrichment in tumors by 4.5-fold. Accordingly, this novel design concept will be inspiring for developing CDT nanomedicines in the future.

One-Pot Rapid Synthesis of Cu2+-Doped GOD@MOF to Amplify the Antitumor Efficacy of Chemodynamic Therapy.

Chemodynamic therapy (CDT) relies on the transformation of intracellular hydrogen peroxide (H2O2) to hydroxyl radicals (·OH) with higher toxicity under the catalysis of Fenton/Fenton-like reagents, which amplifies the oxidative stress and induces significant cellular apoptosis. However, the CDT efficacy is generally limited by the overexpressed GSH and insufficient endogenous H2O2 in tumors. Co-delivery of Cu2+ and glucose oxidase (GOD) can lead to a Cu2+/Cu+ circulation to realize GSH depletion and amplify the Fenton-like reaction. pH-responsive metal-organic frameworks (MOFs) are the optical choice to deliver Fenton/Fenton-like ions to tumors. However, considering that the aqueous condition is requisite for GOD encapsulation, it is challenging to abundantly dope Cu2+ in ZIF-8 MOF nanoparticles in aqueous conditions due to the ease of precipitation and enlarged crystal size. In this work, a robust one-pot biomimetic mineralization method using excessive ligand precursors in aqueous conditions is developed to synthesize GOD@Cu-ZIF-8. Copper ions abundantly doped to the GOD@Cu-ZIF-8 can eliminate GSH to produce Cu+, which is further proceeded to the Fenton-like reaction in the presence of GOD-catalyzed H2O2. Through breaking the tumor microenvironment homeostasis and producing an enhanced CDT effect, the promising antitumor capability of GOD@Cu-ZIF-8 was evidenced by the experiments both in vitro and in vivo.

The Impact and Determinants of Mountainous Topographical Factors on Soil Microbial Community Characteristics.

Soil bacterial and fungal community communities play significant ecological functions in mountain ecosystems. However, it is not clear how topographic factors and soil physicochemical properties influence changes in microbial community structure and diversity. This study aims to investigate how altitude and slope orientation affect soil physicochemical properties, soil microbial communities, and their contributing factors. The assessment was conducted using Illumina MiSeq sequencing in various altitude gradients and on slopes with different aspects (shady slopes and sunny slopes) in the subalpine meadow of Dongling Mountain, Beijing. Topographical factors had a significant effect on soil physicochemical properties: the primary factors determining the structure of microbial communities are total potassium (TK), ammonium nitrogen (NH4+-N), and soil organic carbon (SOC). There was no significant change in the diversity of the bacterial community, whereas the diversity of the fungal community displayed a single-peaked trend. The effect of slope orientation on microbial communities was not as significant as the effect of elevation on them. The number of bacterial communities with significant differences showed a unimodal trend, while the number of fungal communities showed a decreasing trend. The co-occurrence network of fungal communities exhibits greater intricacy than that of bacterial communities, and bacterial communities are more complex in soils with sunny slopes compared to soils with shady slopes, and the opposite is true for fungal communities. The identification of the main factors that control soil microbial diversity and composition in this study, provided the groundwork for investigating the soil microbial response and adaptation to environmental changes in subalpine meadows.

DeepIII: Predicting Isoform-Isoform Interactions by Deep Neural Networks and Data Fusion.

Alternative splicing enables a gene translating into different isoforms and into the corresponding proteoforms, which actually accomplish various biological functions of a living body. Isoform-isoform interactions (IIIs) provide a higher resolution interactome to explore the cellular processes and disease mechanisms than the canonically studied protein-protein interactions (PPIs), which are often recorded at the coarse gene level. The knowledge of IIIs is critical to map pathways, understand protein complexity and functional diversity, but the known IIIs are very scanty. In this paper, we propose a deep learning based method called DeepIII to systematically predict genome-wide IIIs by integrating diverse data sources, including RNA-seq datasets of different human tissues, exon array data, domain-domain interactions (DDIs) of proteins, nucleotide sequences and amino acid sequences. Particularly, DeepIII fuses these data to learn the representation of isoform pairs with a four-layer deep neural networks, and then performs binary classification on the learnt representation to achieve the prediction of IIIs. Experimental results show that DeepIII achieves a superior prediction performance to the state-of-the-art solutions and the III network constructed by DeepIII gives more accurate isoform function prediction. Case studies further confirm that DeepIII can differentiate the individual interaction partners of different isoforms spliced from the same gene. The code and datasets of DeepIII are available at http://mlda.swu.edu.cn/codes.php?name=DeepIII.

Simple sequence-based kernels do not predict protein-protein interactions.

MOTIVATION: A number of methods have been reported that predict protein-protein interactions (PPIs) with high accuracy using only simple sequence-based features such as amino acid 3mer content. This is surprising, given that many protein interactions have high specificity that depends on detailed atomic recognition between physiochemically complementary surfaces. Are the reported high accuracies realistic? RESULTS: We find that the reported accuracies of the predictions are significantly over-estimated, and strongly dependent on the structure of the training and testing datasets used. The choice of which protein pairs are deemed as non-interactions in the training data has a variable impact on the accuracy estimates, and the accuracies can be artificially inflated by a bias towards dominant samples in the positive data which result from the presence of hub proteins in the protein interaction network. To address this bias, we propose a positive set-specific method to create a 'balanced' negative set maintaining the degree distribution for each protein, leading to the conclusion that simple sequence-based features contain insufficient information to be useful for predicting PPIs, but that protein domain-based features have some predictive value. AVAILABILITY: Our method, named 'BRS-nonint', is available at http://www.bioinformatics.leeds.ac.uk/BRS-nonint/. All the datasets used in this study are derived from publicly available data, and are available at http://www.bioinformatics.leeds.ac.uk/BRS-nonint/PPI_RandomBalance.html CONTACT: maozuguo@hit.edu.cn; d.r.westhead@leeds.ac.uk.