Multi-omics network characterization reveals novel microRNA biomarkers and mechanisms for diagnosis and subtyping of kidney transplant rejection

Background Kidney transplantation is an optimal method for treatment of end-stage kidney failure. However, kidney transplant rejection (KTR) is commonly observed to have negative effects on allograft function. MicroRNAs (miRNAs) are small non-coding RNAs with regulatory role in KTR genesis, the identification of miRNA biomarkers for accurate diagnosis and subtyping of KTR is therefore of clinical significance for active intervention and personalized therapy. Methods In this study, an integrative bioinformatics model was developed based on multi-omics network characterization for miRNA biomarker discovery in KTR. Compared with existed methods, the topological importance of miRNA targets was prioritized based on cross-level miRNA-mRNA and protein–protein interaction network analyses. The biomarker potential of identified miRNAs was computationally validated and explored by receiver-operating characteristic (ROC) evaluation and integrated “miRNA-gene-pathway” pathogenic survey. Results Three miRNAs, i.e., miR-145-5p, miR-155-5p, and miR-23b-3p, were screened as putative biomarkers for KTR monitoring. Among them, miR-155-5p was a previously reported signature in KTR, whereas the remaining two were novel candidates both for KTR diagnosis and subtyping. The ROC analysis convinced the power of identified miRNAs as single and combined biomarkers for KTR prediction in kidney tissue and blood samples. Functional analyses, including the latent crosstalk among HLA-related genes, immune signaling pathways and identified miRNAs, provided new insights of these miRNAs in KTR pathogenesis. Conclusions A network-based bioinformatics approach was proposed and applied to identify candidate miRNA biomarkers for KTR study. Biological and clinical validations are further needed for translational applications of the findings.

The Use of 3D Convolutional Autoencoder in Fault and Fracture Network Characterization

Conventional pattern recognition methods directly use 1D poststack data or 2D prestack data for the statistical pattern recognition of fault and fracture network, thereby ignoring the spatial structure information in 3D seismic data. As a result, the generated fault and fracture network is not distinguishable and has poor continuity. In this paper, a fault and fracture network characterization method based on 3D convolutional autoencoder is proposed. First, in the autoencoder training frame, 3D prestack data are used as input, and the 3D convolution operation is used to mine the spatial structure information to the maximum and gradually reduce the spatial dimension of the input. Then, the residual network is used to recover the input’s details and the corresponding spatial dimension. Lastly, the hidden features extracted by the encoders are recognized via k -means, SOM, and two-step clustering analysis. The validity of the method is verified by testing the seismic simulation data and applying real seismic data. The 3D convolution can directly process the seismic data and maximize the prestack texture attributes and spatial structure information provided by 3D seismic data without dimensionality reduction and other preprocessing operations. The interleaving convolution layer and residual block overcome low learning and accuracy rates due to the deepening of networks.