Search

Motivation Multiple sequence alignment (MSA) is one of the hotspots of current research and is commonly used in sequence analysis scenarios. However, there is no lasting solution for MSA because it is a Nondeterministic Polynomially complete problem, and the existing methods still have room to improve the accuracy. Results We propose Deep reinforcement learning with Positional encoding and self-Attention for MSA, based on deep reinforcement learning, to enhance the accuracy of the alignment Specifically, inspired by the translation technique in natural language processing, we introduce self-attention and positional encoding to improve accuracy and reliability. Firstly, positional encoding encodes the position of the sequence to prevent the loss of nucleotide position information. Secondly, the self-attention model is used to extract the key features of the sequence. Then input the features into a multi-layer perceptron, which can calculate the insertion position of the gap according to the features. In addition, a novel reinforcement learning environment is designed to convert the classic progressive alignment into progressive column alignment, gradually generating each column's sub-alignment. Finally, merge the sub-alignment into the complete alignment. Extensive experiments based on several datasets validate our method's effectiveness for MSA, outperforming some state-of-the-art methods in terms of the Sum-of-pairs and Column scores. Availability and implementation The process is implemented in Python and available as open-source software from https://github.com/ZhangLab312/DPAMSA. [ABSTRACT FROM AUTHOR]

Enhancers are crucial cis- regulatory elements that control gene expression in a cell-type-specific manner. Despite extensive genetic and computational studies, accurately predicting enhancer activity in different cell types remains a challenge, and the grammar of enhancers is still poorly understood. Here, we present HEAP (h igh-resolution e nhancer a ctivity p rediction), an explainable deep learning framework for predicting enhancers and exploring enhancer grammar. The framework includes three modules that use grammar-based reasoning for enhancer prediction. The algorithm can incorporate DNA sequences and epigenetic modifications to obtain better accuracy. We use a novel two-step multi-task learning method, task adaptive parameter sharing (TAPS), to efficiently predict enhancers in different cell types. We first train a shared model with all cell-type datasets. Then we adapt to specific tasks by adding several task-specific subset layers. Experiments demonstrate that HEAP outperforms published methods and showcases the effectiveness of the TAPS, especially for those with limited training samples. Notably, the explainable framework HEAP utilizes post-hoc interpretation to provide insights into the prediction mechanisms from three perspectives: data, model architecture and algorithm, leading to a better understanding of model decisions and enhancer grammar. To the best of our knowledge, HEAP will be a valuable tool for insight into the complex mechanisms of enhancer activity. [ABSTRACT FROM AUTHOR]

Antibiotic-resistant bacteria are current threats to available antibiotic therapies, and this has renewed interest in the therapeutic use of phage as an alternative. However, development of phage resistance has led to unsuccessful therapeutic outcomes. In the current study, we applied phage training to minimize bacterial phage resistance and to improve treatment outcome by adapting the phage to their target hosts during co-evolution. We isolated and characterized a novel Pseudomonas aeruginosa N4-like lytic phage (PWJ) from wastewater in Yangzhou, China. PWJ is a double-stranded DNA podovirus that can efficiently lyse the model strain ATCC 27,853 and opportunistic pathogen PAO1. Genome sequencing of PWJ revealed features similar to those of the N4-like P. aeruginosa phage YH6. We used PWJ to screen for an evolved trained phage (WJ_Ev14) that restored infectivity to PWJ phage bacterial resisters. BLASTN analysis revealed that WJ_Ev14 is identical to its ancestor PWJ except for the amino acid substitution R1051S in its tail fiber protein. Moreover, phage adsorption tests and transmission electron microscopy of resistant bacteria demonstrated that the R1051S substitution was most likely the reason WJ_Ev14 could re-adsorb and regain infectivity. Furthermore, phage therapy assays in vitro and in a mouse P. aeruginosa lung infection model demonstrated that PWJ treatment resulted in improved clinical results and a reduction in lung bacterial load whereas the joint phage cocktail (PWJ+ WJ_Ev14) was better able to delay the emergence of resister bacteria. The phage cocktail (PWJ +WJ_Ev14) represents a promising candidate for inclusion in phage cocktails developed for clinical applications. [ABSTRACT FROM AUTHOR]

To clarify the role of tree characteristics and slope positions in the time lag between tree stem sap flux density (J s) and solar radiation (R s). Plants of different diameter classes in a Larix olgensis near-mature forest (31 years old) in the hilly area of the Sanjiang Plain were used. The relationships between the time lag J s– R s and tree characteristics, adjacent tree characteristics and slope positions were evaluated. Though both J s and R s exhibited diurnal variation, they were not synchronized, thus leading to a time lag between J s and R s. During the growing season, the change in J s lagged behind the change in R s by 21.1 ± 6.9 min. Compared with tree height and crown width, the time lag J s– R s was more dependent on diameter at breast height (DBH). The time lag between J s and R s showed a linear increase with DBH. Compared with the characteristics of neighboring trees, the time lags J s– R s were more dependent on their own tree characteristics. A significant relationship was not observed between the time lag J s– R s and soil volumetric water content. The effects of tree characteristics, adjacent tree characteristics and slope positions on the formation of the time lag J s– R s were compared. The time lag of J s on R s was mainly controlled by the tree characteristics (DBH). DBH is an important factor that affects the time lag between J s and R s under sunny conditions during the growing season of L. olgensis. [ABSTRACT FROM AUTHOR]

Motivation Transcription factor (TF) binds to conservative DNA binding sites in different cellular environments and development stages by physical interaction with interdependent nucleotides. However, systematic computational characterization of the relationship between higher-order nucleotide dependency and TF-DNA binding mechanism in diverse cell types remains challenging. Results Here, we propose a novel multi-task learning framework HAMPLE to simultaneously predict TF binding sites (TFBS) in distinct cell types by characterizing higher-order nucleotide dependencies. Specifically, HAMPLE first represents a DNA sequence through three higher-order nucleotide dependencies, including k -mer encoding, DNA shape and histone modification. Then, HAMPLE uses the customized gate control and the channel attention convolutional architecture to further capture cell-type-specific and cell-type-shared DNA binding motifs and epigenomic languages. Finally, HAMPLE exploits the joint loss function to optimize the TFBS prediction for different cell types in an end-to-end manner. Extensive experimental results on seven datasets demonstrate that HAMPLE significantly outperforms the state-of-the-art approaches in terms of auROC. In addition, feature importance analysis illustrates that k -mer encoding, DNA shape, and histone modification have predictive power for TF-DNA binding in different cellular environments and are complementary to each other. Furthermore, ablation study, and interpretable analysis validate the effectiveness of the customized gate control and the channel attention convolutional architecture in characterizing higher-order nucleotide dependencies. Availability and implementation The source code is available at https://github.com/ZhangLab312/Hample. [ABSTRACT FROM AUTHOR]

Aims To evaluate the effects of the Qingwen Gupi decoction (QGT) in a rat model of bleomycin-induced pulmonary fibrosis (PF), and explore the underlying mechanisms by integrating UPLC-Q-TOF/MS metabolomics and 16S rDNA sequencing of gut microbiota. Methods and Results The animals were randomly divided into the control, PF model, pirfenidone-treated, and low-, medium-, and high-dose QGT groups. The lung tissues were examined and the expression of TGF-β, SMAD-3, and SMAD-7 mRNAs in the lung tissues were analyzed. Metabolomic profiles were analyzed by UPLC-QTOF/MS, and the intestinal flora were examined by prokaryotic 16 rDNA sequencing. Pathological examination and biochemical indices revealed that QGT treatment improved the symptoms of PF by varying degrees. Furthermore, QGT significantly downregulated TGF-β1 and Smad-3 mRNAs and increased the expression levels of Smad-7. QGT-L in particular increased the levels of 18 key metabolic biomarkers that were associated with nine gut microbial species and may exert antifibrosis effects through arachidonic acid metabolism, glycerophospholipid metabolism, and phenylalanine metabolism. Conclusions QGT alleviated PF in a rat model through its anti-inflammatory, antioxidant, and anti-fibrotic effects, and by reversing bleomycin-induced gut dysbiosis.This study lays the foundation for further research on the pathological mechanisms of PF and the development of new drug candidates. [ABSTRACT FROM AUTHOR]

The discovery of putative transcription factor binding sites (TFBSs) is important for understanding the underlying binding mechanism and cellular functions. Recently, many computational methods have been proposed to jointly account for DNA sequence and shape properties in TFBSs prediction. However, these methods fail to fully utilize the latent features derived from both sequence and shape profiles and have limitation in interpretability and knowledge discovery. To this end, we present a novel Deep Convolution Attention network combining Sequence and Shape, dubbed as D-SSCA, for precisely predicting putative TFBSs. Experiments conducted on 165 ENCODE ChIP-seq datasets reveal that D-SSCA significantly outperforms several state-of-the-art methods in predicting TFBSs, and justify the utility of channel attention module for feature refinements. Besides, the thorough analysis about the contribution of five shapes to TFBSs prediction demonstrates that shape features can improve the predictive power for transcription factors-DNA binding. Furthermore, D-SSCA can realize the cross-cell line prediction of TFBSs, indicating the occupancy of common interplay patterns concerning both sequence and shape across various cell lines. The source code of D-SSCA can be found at https://github.com/MoonLord0525/. [ABSTRACT FROM AUTHOR]

Transcription factors (TFs) are essential proteins in regulating the spatiotemporal expression of genes. It is crucial to infer the potential transcription factor binding sites (TFBSs) with high resolution to promote biology and realize precision medicine. Recently, deep learning-based models have shown exemplary performance in the prediction of TFBSs at the base-pair level. However, the previous models fail to integrate nucleotide position information and semantic information without noisy responses. Thus, there is still room for improvement. Moreover, both the inner mechanism and prediction results of these models are challenging to interpret. To this end, the D eep A ttentive E ncoder- D ecoder Neural Net work (D-AEDNet) is developed to identify the location of TFs–DNA binding sites in DNA sequences. In particular, our model adopts Skip Architecture to leverage the nucleotide position information in the encoder and removes noisy responses in the information fusion process by Attention Gate. Simultaneously, the T ranscription F actor M otif D iscovery based on S liding W indow (TF-MoDSW), an approach to discover TFs–DNA binding motifs by utilizing the output of neural networks, is proposed to understand the biological meaning of the predicted result. On ChIP-exo datasets, experimental results show that D-AEDNet has better performance than competing methods. Besides, we authenticate that Attention Gate can improve the interpretability of our model by ways of visualization analysis. Furthermore, we confirm that ability of D-AEDNet to learn TFs–DNA binding motifs outperform the state-of-the-art methods and availability of TF-MoDSW to discover biological sequence motifs in TFs–DNA interaction by conducting experiment on ChIP-seq datasets. [ABSTRACT FROM AUTHOR]

Long awns are important for seed dispersal in wild rice (Oryza rufipogon), but are absent in cultivated rice (Oryza sativa). The genetic mechanism involved in loss-of-awn in cultivated rice remains unknown. We report here the molecular cloning of a major quantitative trait locus, An-1, which regulates long awn formation in O. rufipogon. An-1 encodes a basic helix-loop-helix protein, which regulates cell division. The nearly-isogenic line (NIL- An-1) carrying a wild allele An-1 in the genetic background of the awnless indica Guangluai4 produces long awns and longer grains, but significantly fewer grains per panicle compared with Guangluai4. Transgenic studies confirmed that An-1 positively regulates awn elongation, but negatively regulates grain number per panicle. Genetic variations in the An-1 locus were found to be associated with awn loss in cultivated rice. Population genetic analysis of wild and cultivated rice showed a significant reduction in nucleotide diversity of the An-1 locus in rice cultivars, suggesting that the An-1 locus was a major target for artificial selection. Thus, we propose that awn loss was favored and strongly selected by humans, as genetic variations at the An-1 locus that cause awn loss would increase grain numbers and subsequently improve grain yield in cultivated rice. [ABSTRACT FROM AUTHOR]

Seed shattering is an important agricultural trait in crop domestication. SH4 (for grain shattering quantitative trait locus on chromosome 4) and qSH1 (for quantitative trait locus of seed shattering on chromosome 1) genes have been identified as required for reduced seed shattering during rice (Oryza sativa) domestication. However, the regulatory pathways of seed shattering in rice remain unknown. Here, we identified a seed shattering abortion1 (shat1) mutant in a wild rice introgression line. The SHAT1 gene, which encodes an APETALA2 transcription factor, is required for seed shattering through specifying abscission zone (AZ) development in rice. Genetic analyses revealed that the expression of SHAT1 in AZ was positively regulated by the trihelix transcription factor SH4. We also identified a frameshift mutant of SH4 that completely eliminated AZs and showed nonshattering. Our results suggest a genetic model in which the persistent and concentrated expression of active SHAT1 and SH4 in the AZ during early spikelet developmental stages is required for conferring AZ identification. qSH1 functioned downstream of SHAT1 and SH4 , through maintaining SHAT1 and SH4 expression in AZ, thus promoting AZ differentiation. [ABSTRACT FROM AUTHOR]

A huge amount of cDNA and EST resources have been developed for cultivated rice species Oryza sativa; however, only few cDNA resources are available for wild rice species. In this study, we isolated and completely sequenced 1888 putative full-length cDNA (FLcDNA) clones from wild rice Oryza rufipogon Griff. W1943 for comparative analysis between wild and cultivated rice species. Two cDNA libraries were constructed from 3-week-old leaf samples under either normal or cold-treated conditions. Homology searching of these cDNA sequences revealed that >96.8% of the wild rice cDNAs were matched to the cultivated rice O. sativa ssp. japonica cv. Nipponbare genome sequence. However, <22% of them were fully matched to the cv. Nipponbare genome sequence. The comparative analysis showed that O. rufipogon W1943 had greater similarity to O. sativa ssp. japonica than to ssp. indica cultivars. In addition, 17 novel rice cDNAs were identified, and 41 putative tissue-specific expression genes were defined through searching the rice massively parallel signature-sequencing database. In conclusion, these FLcDNA clones are a resource for further function verification and could be broadly utilized in rice biological studies. [ABSTRACT FROM PUBLISHER]