PNNGS, a multi-convolutional parallel neural network for genomic selection.

Genomic selection (GS) can accomplish breeding faster than phenotypic selection. Improving prediction accuracy is the key to promoting GS. To improve the GS prediction accuracy and stability, we introduce parallel convolution to deep learning for GS and call it a parallel neural network for genomic selection (PNNGS). In PNNGS, information passes through convolutions of different kernel sizes in parallel. The convolutions in each branch are connected with residuals. Four different Lp loss functions train PNNGS. Through experiments, the optimal number of parallel paths for rice, sunflower, wheat, and maize is found to be 4, 6, 4, and 3, respectively. Phenotype prediction is performed on 24 cases through ridge-regression best linear unbiased prediction (RRBLUP), random forests (RF), support vector regression (SVR), deep neural network genomic prediction (DNNGP), and PNNGS. Serial DNNGP and parallel PNNGS outperform the other three algorithms. On average, PNNGS prediction accuracy is 0.031 larger than DNNGP prediction accuracy, indicating that parallelism can improve the GS model. Plants are divided into clusters through principal component analysis (PCA) and K-means clustering algorithms. The sample sizes of different clusters vary greatly, indicating that this is unbalanced data. Through stratified sampling, the prediction stability and accuracy of PNNGS are improved. When the training samples are reduced in small clusters, the prediction accuracy of PNNGS decreases significantly. Increasing the sample size of small clusters is critical to improving the prediction accuracy of GS.

RASSF1A promotes ATM signaling and RASSF1A methylation is a synthetic lethal marker for ATR inhibitors.

Aim: The mechanism of RASSF1A in DNA damage repair remains to be further clarified for applying to synthetic lethal strategy. Materials & methods: Eight esophageal cancer cell lines, 181 cases of esophageal dysplasia and 1066 cases of primary esophageal squamous cell carcinoma (ESCC) were employed. Methylation-specific PCR, the CRISPR/Cas9 technique, immunoprecipitation assay and a xenograft mouse model were used. Results: RASSF1A was methylated in 2.21% of esophageal dysplasia and 11.73% of ESCC. RASSF1A was also involved in DNA damage repair through activating Hippo signaling. Loss of RASSF1A expression sensitized esophageal cancer cell lines to ataxia telangiectasia mutated and rad3-related (ATR) inhibitor (VE-822) both in vitro and in vivo. Conclusion: RASSF1A methylation is a synthetic lethal marker for ATR inhibitors.

Dynamic transcriptome and network-based analysis of yellow leaf mutant Ginkgo biloba.

BACKGROUND: Golden leaf in autumn is a prominent feature of deciduous tree species like Ginkgo biloba L., a landscape tree widely cultivated worldwide. However, little was known about the molecular mechanisms of leaf yellowing, especially its dynamic regulatory network. Here, we performed a suite of comparative physiological and dynamic transcriptional analyses on the golden-leaf cultivar and the wild type (WT) ginkgo to investigate the underlying mechanisms of leaf yellowing across different seasons. RESULTS: In the present study, we used the natural bud mutant cultivar with yellow leaves "Wannianjin" (YL) as materials. Physiological analysis revealed that higher ratios of chlorophyll a to chlorophyll b and carotenoid to chlorophyll b caused the leaf yellowing of YL. On the other hand, dynamic transcriptome analyses showed that genes related to chlorophyll metabolism played key a role in leaf coloration. Genes encoding non-yellow coloring 1 (NYC1), NYC1-like (NOL), and chlorophyllase (CLH) involved in the degradation of chlorophyll were up-regulated in spring. At the summer stage, down-regulated HEMA encoding glutamyl-tRNA reductase functioned in chlorophyll biosynthesis, while CLH involved in chlorophyll degradation was up-regulated, causing a lower chlorophyll accumulation. In carotenoid metabolism, genes encoding zeaxanthin epoxidase (ZEP) and 9-cis-epoxy carotenoid dioxygenase (NCED) showed significantly different expression levels in the WT and YL. Moreover, the weighted gene co-expression network analysis (WGCNA) suggested that the most associated transcriptional factor, which belongs to the AP2/ERF-ERF family, was engaged in regulating pigment metabolism. Furthermore, quantitative experiments validated the above results. CONCLUSIONS: By comparing the golden-leaf cultivar and the wide type of ginkgo across three seasons, this study not only confirm the vital role of chlorophyll in leaf coloration of YL but also provided new insights into the seasonal transcriptome landscape and co-expression network. Our novel results pinpoint candidate genes for further wet-bench experiments in tree species.

Temporal Dynamic Transcriptome Landscape Reveals Regulatory Network During the Early Differentiation of Female Strobilus Buds in Ginkgo biloba.

Reproductive bud differentiation is one of the most critical events for the reproductive success of seed plants. Yet, our understanding of genetic basis remains limited for the development of the reproductive organ of gymnosperms, namely, unisexual strobilus or cone, leaving its regulatory network largely unknown for strobilus bud differentiation. In this study, we analyzed the temporal dynamic landscapes of genes, long non-coding RNAs (lncRNAs), and microRNAs (miRNAs) during the early differentiation of female strobilus buds in Ginkgo biloba based on the whole transcriptome sequencing. Results suggested that the functions of three genes, i.e., Gb_19790 (GbFT), Gb_13989 (GinNdly), and Gb_16301 (AG), were conserved in both angiosperms and gymnosperms at the initial differentiation stage. The expression of genes, lncRNAs, and miRNAs underwent substantial changes from the initial differentiation to the enlargement of ovule stalk primordia. Besides protein-coding genes, 364 lncRNAs and 15 miRNAs were determined to be functional. Moreover, a competing endogenous RNA (ceRNA) network comprising 10,248 lncRNA-miRNA-mRNA pairs was identified, which was highly correlated with the development of ovulate stalk primordia. Using the living fossil ginkgo as the study system, this study not only reveals the expression patterns of genes related to flowering but also provides novel insights into the regulatory networks of lncRNAs and miRNAs, especially the ceRNA network, paving the way for future studies concerning the underlying regulation mechanisms of strobilus bud differentiation.

TOD-CUP: a gene expression rank-based majority vote algorithm for tissue origin diagnosis of cancers of unknown primary.

Gene expression profiling holds great potential as a new approach to histological diagnosis and precision medicine of cancers of unknown primary (CUP). Batch effects and different data types greatly decrease the predictive performance of biomarker-based algorithms, and few methods have been widely applied to identify tissue origin of CUP up to now. To address this problem and assist in more precise diagnosis, we have developed a gene expression rank-based majority vote algorithm for tissue origin diagnosis of CUP (TOD-CUP) of most common cancer types. Based on massive tissue-specific RNA-seq data sets (10 553) found in The Cancer Genome Atlas (TCGA), 538 feature genes (biomarkers) were selected based on their gene expression ranks and used to predict tissue types. The top scoring pairs (TSPs) classifier of the tumor type was optimized by the TCGA training samples. To test the prediction accuracy of our TOD-CUP algorithm, we analyzed (1) two microarray data sets (1029 Agilent and 2277 Affymetrix/Illumina chips) and found 91% and 94% prediction accuracy, respectively, (2) RNA-seq data from five cancer types derived from 141 public metastatic cancer tumor samples and achieved 94% accuracy and (3) a total of 25 clinical cancer samples (including 14 metastatic cancer samples) were able to classify 24/25 samples correctly (96.0% accuracy). Taken together, the TOD-CUP algorithm provides a powerful and robust means to accurately identify the tissue origin of 24 cancer types across different data platforms. To make the TOD-CUP algorithm easily accessible for clinical application, we established a Web-based server for tumor tissue origin diagnosis (http://ibi. zju.edu.cn/todcup/).

Characteristics of plant circular RNAs.

Circular RNA (circRNA) is a kind of covalently closed single-stranded RNA molecules that have been proved to play important roles in transcriptional regulation of genes in diverse species. With the rapid development of bioinformatics tools, a huge number (95143) of circRNAs have been identified from different plant species, providing an opportunity for uncovering the overall characteristics of plant circRNAs. Here, based on publicly available circRNAs, we comprehensively analyzed characteristics of plant circRNAs with the help of various bioinformatics tools as well as in-house scripts and workflows, including the percentage of coding genes generating circRNAs, the frequency of alternative splicing events of circRNAs, the non-canonical splicing signals of circRNAs and the networks involving circRNAs, miRNAs and mRNAs. All this information has been integrated into an upgraded online database, PlantcircBase 3.0 (http://ibi.zju.edu.cn/plantcircbase/). In this database, we provided browse, search and visualization tools as well as a web-based blast tool, BLASTcirc, for prediction of circRNAs from query sequences based on searching against plant genomes and transcriptomes.