155 results about "Multi omics" patented technology

The invention provides a cross-omics sparse feature selection system and method based on hierarchical causal modeling, and the system comprises a data input and preprocessing module which is used for receiving multi-omics original data of a multivariate sample; the hierarchical causal structure learning module is connected with the data input and adaptive preprocessing module and is used for constructing a cross-omics hierarchical causal topology; the causal-oriented sparse feature selection module is connected with the hierarchical causal structure learning module; and the model retraining and integration module is used for constructing a three-layer weighted integration discrimination model based on the screened markers, optimizing the fusion weight of each layer through a gradient descent algorithm, and outputting a final prediction result. According to the method, the protein-metabolism biological hierarchy relationship and serum-urine complementary information are fully utilized, and the method has good generalization ability and can be widely applied to marker mining and prediction modeling of cancers, metabolic diseases and the like, so that the accuracy and reliability of precise medical treatment are improved.

The invention discloses a multi-omics cancer subtype recognition method, system and device based on density sensing cluster structure guide contrast learning and a medium, and belongs to the technical field of bioinformatics and artificial intelligence crossing. The method comprises the following steps: acquiring and preprocessing multi-omics data; constructing an omics specific auto-encoder and learning potential representation; constructing a density sensing cluster block in the potential space; constructing a cross-omics positive and negative sample pair based on cluster block sample overlapping; difficult negative sample mining; constructing a cluster block level cross-omics contrast learning target, and training and updating; a self-supervised soft refinement mechanism is introduced to dynamically enhance a cluster structure; and carrying out multi-loss joint optimization and model iteration training. According to the method, the robustness and the stability of a cancer subtype recognition result can be improved, high-dimensional, multi-source and multi-noise multi-omics data can be efficiently modeled and analyzed, good generalization ability and application potential are achieved, and reliable technical support can be provided for cancer subtype research, patient stratified analysis and precise medical aid decision making.

The invention discloses a multi-omics data integration and classification method, system and device based on hierarchical attention, and is applied to the field of precise medical big data analysis. The method comprises the following steps: firstly, generating feature embedding and feature importance scores through a plurality of parallel feature-level attention modules; then, embedding and inputting all the characteristics of the omics into a unified omics-level attention module, and generating omics embedding and omics importance scores; and finally, a classification prediction task is executed based on omics embedding, and a classification result is output for disease classification. The invention completely abandons a traditional dependency graph convolutional network and an integration normal form of variants of the dependency graph convolutional network, and provides a universal hierarchical attention integration architecture. The framework supports classification tasks of any complex diseases, is not limited by omics data types and combination modes, not only is remarkably superior to a traditional integration normal form in classification performance, but also shows a unique negative generalization distance, and proves that the framework has excellent generalization ability. Meanwhile, features and omics importance scores automatically output by the model provide a powerful analysis tool for biomarker discovery and precise diagnosis and treatment of complex diseases.

The invention belongs to the technical field of gene identification, and discloses a method for integrating multi-omics data to enhance genome prediction and candidate gene identification, candidate gene identification is verified through multi-layer evidence integration, and the verification comprises priority ordering based on gene contribution scores, function enrichment analysis, generic genome network verification and CRISPR / Cas9 experimental verification. Evaluation on a corn population (n = 174) containing complete genomics, transcriptomics, translational omics and proteomics maps shows that the framework is remarkably improved in grain character prediction and is improved by 2.9-12.3% compared with a genome selection baseline, and meanwhile candidate genes verified by experiments are recognized. The invention further verifies the universality of the framework to five traits on an arabidopsis thaliana population, and provides an open source software platform to promote the practical application of the framework in a breeding plan.

Provided herein are systems, devices, and methods for processing, analyzing, and classifying biological data sets and generation of cell profiles. The data sets may include multi-omic data. Some embodiments may include the use of machine learning in training a classifier of raw multi-omic data and incorporating system biology knowledge to understand cellular behavior and cell status at the biomolecular level.

Described herein are methods such as multi-omic methods for assessing a disease such as cancer. The multi-omic methods may integrate proteomic, transcriptomic, genomic, lipidomic, or metabolomic data. The method screening diseases or disease states. Also described herein are methods for screening for diseases or disease states from biological samples. The methods may include assessing whether a nodule, mass, or cyst is cancerous.

The invention discloses a multi-layer heterogeneous network unicellular organism network inference method based on meta-path enhancement, which mainly comprises a gene regulation knowledge base enhanced multi-layer heterogeneous network construction module for integrating an external gene interaction network and multiple omics data such as scRNA-seq, scATAC-seq, ST and the like; constructing a single-cell multi-omics multilayer heterogeneous network containing cell-cell, cell-gene and gene-gene relationships, and fusing spatial constraints to consider cell positions and tissue structures; and the feature enhancement module based on the meta-path explores complex semantics of the network by designing a multi-hop meta-path mode, designs an adaptive multi-view learning framework and a multi-round enhancement mechanism, and optimizes feature representation by using cell-gene interaction and cross-modal attention fusion. The unicellular biological network can be effectively deduced, the deduction accuracy and biological interpretation are remarkably improved, the method plays an important role in understanding the cell biological process, developing and treating diseases and the like, has good expandability, and can further integrate multi-modal omics data such as proteomics and metabonomics.

The invention discloses a hepatocellular carcinoma prognosis and immune response prediction method based on multi-omics machine learning, and relates to the technical field of biomedicine and artificial intelligence crossing, and the method comprises the following steps: S1, constructing a multi-omics data set for model training and verification; s2, performing feature integration and clustering analysis on the multi-omics data to obtain corresponding hepatocellular carcinoma molecular subtype distribution; s3, constructing a hepatocellular carcinoma prognosis model based on Cox regression combined with a random survival forest, identifying 11 core immune genes and corresponding weight coefficients by training the model, and constructing an immunotherapy response index IMLIRI score; and S4, carrying out clinical application on the IMLIRI score. According to the method, through multi-omics data integration and multi-queue external verification, the influence of data deviation and queue heterogeneity on the model performance is reduced, so that the method shows stable prediction performance in hepatocellular carcinoma queues with different sources and different pathogenesis backgrounds, and the reliability and generalizability of the model in clinical application are improved.

The invention relates to a disease-specific quantitative trait locus identification method based on multi-omics integration. The method comprises the following steps: acquiring variation sites of whole genome sequencing data of a target object, and molecular phenotypes and molecular abundance of molecular phenotype data; determining an association significance probability value of an association pair formed by the variation point and the molecular phenotype based on the variation point and the molecular abundance, and screening a first association pair from the association pair based on the association significance probability value and condition analysis; determining a consistent second association pair in the normal association pair and the disease association pair, and determining a third association pair with a disease interaction effect in the second association pair; calculating a first effect estimation value and a second effect estimation value of each third association pair; and based on the first effect estimation value and the second effect estimation value of the third correlation pair, determining a target correlation pair related to the Parkinson's disease, and taking the target correlation pair as the identified quantitative trait site. By adopting the method, the Parkinson's specific pathogenic heritable variation can be accurately identified.

The application provides a miRNA-disease relationship prediction method, system and model based on a hybrid expert model and a storage medium. The method comprises the following steps: obtaining a miRNA-disease correlation matrix of multi-omics data of a miRNA-disease to be predicted, a miRNA similarity matrix and a disease similarity matrix; constructing a miRNA-disease heterogeneous graph, a miRNA homogeneous subgraph and a disease homogeneous subgraph, and fusing them into a multi-modal biological graph network; inputting the fusion network into a pre-trained gated hybrid multi-expert network model to output a correlation probability of a miRNA-disease pair to be predicted; and the gated hybrid multi-expert network model comprises multiple expert networks, a gating network and an output layer. The application adopts a gating mechanism to adjust the weights of each expert for miRNA-disease prediction, maintains high prediction accuracy on a small-scale data set, and has the advantages of portability and light weight.

The present application belongs to the field of biotechnology and genomics, and particularly relates to a method and system for determining the chromosome base number of Macrobrachium rosenbergii based on multi-omics joint analysis. The method obtains de novo assembly sequencing data and Hi-C sequencing data, generates a chromosome-level candidate assembly without presetting the number of chromosomes by using Hi-C interaction signals after primary assembly, and performs whole-genome collinearity alignment with no less than two published reference genomes; in combination with quality constraints such as collinearity continuity, Hi-C boundary characteristics and BUSCO / LAI, the candidate chromosome boundary is comprehensively judged and iteratively converged, and finally the chromosome base number and reviewable evidence chain are output. The embodiments show that the present application can identify and correct the number redundancy caused by over-splitting of the reference genome, determine the base number of Macrobrachium rosenbergii as n=57 (2n=114), and improve the objectivity and reliability of base number determination.

The invention relates to the technical field of image processing, and further relates to a tumor characteristic spectrum construction method based on CTC enrichment and multi-omics analysis. The method comprises the following steps: 1, sequentially carrying out spiral inertial focusing and deterministic lateral displacement sorting on a whole blood sample to be detected, and finishing single-cell microcavity positioning capture of circulating tumor cells in a single-cell microcavity capture array region; 2, acquiring a time-varying curve of an oxygen molecule concentration value for each positioned and captured circulating tumor cell, and acquiring a metabolic function characteristic parameter group; 3, forming a multi-omics joint feature vector in combination with the metabolic function feature parameter group; and 4, constructing a circulating tumor cell subset based on the multi-omics combined feature vector, extracting symbolic molecular features, and integrating and outputting a tumor feature spectrum. According to the invention, more stable technical support can be provided for tumor heterogeneity analysis, metastasis potential evaluation and curative effect monitoring.

The invention relates to the technical field of biological information, and particularly discloses a multi-omics genome combined genetic assessment method and device. According to the method, multi-source data such as a genome, a transcriptome, a proteome, a metabolome, a microbiome and a high-throughput phenotype are integrated by constructing a multi-omics fusion similar matrix, an adaptive dichotomy is adopted to optimize omics weights, a grid search and cross validation system is introduced, and global optimal estimation of parameters is achieved. Verification of the method on livestock and poultry breeding and plant breeding data shows that the prediction accuracy of complex characters can be remarkably improved, and particularly, the prediction precision in low heritability characters is remarkably improved. The invention further provides a corresponding device which comprises a data acquisition module, a matrix construction module, a parameter optimization module and a breeding value calculation module, comprehensive utilization of multi-omics information can be efficiently achieved, and the accuracy and calculation efficiency of breeding value estimation are improved.

Microscopy imaging that allows for multiple mRNAs, proteins and metabolites to be spatially resolved at a subcellular level provides valuable molecular information which is a crucial factor for understanding tissue heterogeneity as for example within the tumor micro environment. The current invention describes a method (High Density-SUMI-Seq) which combines the use of Spatial Unique Molecular Identifier in situ localization and identification (by in situ sequencing or sequential fluorescence hybridization) of rolonies derived from rolling circle amplification of circular oligonucleotides and in vitro sequencing of target amplified RNA or DNA in combination with SUMI identification at a subcellular level with no optical diffraction limitation in the amount of amplified target information that can be analyzed per cell. Apart from amplified RNA or DNA, the High Density-SUMI-Seq method can also be applied using linear oligonucleotides to spatially resolve proteins and metabolites to provide multiomics results.

The invention discloses a spatial multi-omics data prediction method and model based on spatial constraint and adversarial learning, and application. The prediction method comprises the steps of modal specific coding and potential variable inference; carrying out spatial dependency modeling based on a sparse variational Gaussian process; modeling potential variables of spatial dependency constraints; carrying out multi-modal potential representation alignment based on adversarial learning; modal specific decoding and cross-modal feature prediction are carried out; and combining objective function construction and model training. According to the embodiment of the invention, space coordinate information is fully utilized to restrain potential representation, an adversarial learning mechanism is introduced to realize effective alignment of potential features of different modals, cross-modal feature prediction is realized on the basis, and the precision, stability and application value of space multi-omics data analysis are improved.

The invention relates to the technical field of plant biology, and discloses a method for improving stress resistance of secondary metabolites of tea trees. The method comprises the following steps: collecting multi-omics time sequence data of tea trees under drought stress, and constructing a time-space aligned multi-dimensional data matrix; modeling a dynamic causal regulation relationship among genes, metabolites and phenotypes by utilizing a dynamic Bayesian network; designing a multi-dimensional evaluation algorithm fusing network topology and biological effect to screen key regulatory genes; the influence on target metabolites such as catechin is quantitatively predicted through in-vivo computer disturbance simulation; and finally generating an optimal gene editing or molecular marker assisted breeding scheme. According to the method, breakthrough from correlation analysis to causal inference is achieved, and the accuracy and efficiency of collaborative improvement of the stress resistance and the secondary metabolites of the tea trees are improved.

The invention provides a multi-view subspace clustering cancer subtype identification method based on self-reinforcement learning, and the method comprises the steps: firstly, extracting the potential feature representation of each view from multi-omics data through a potential feature learning module; then, clustering similar samples by using a self-expression learning module, and introducing initial graph information as a supervision signal to construct a self-expression coefficient matrix; secondly, inputting the matrix into a view image fusion unit, and fusing multi-view information to generate a consensus image; in addition, in order to further suppress noise interference in multi-omics data, a self-strengthening back propagation unit is introduced, a confidence matrix is generated by optimizing a self-expression coefficient, fusion loss back propagation is guided, the quality of the self-expression coefficient is iteratively improved, and a consensus graph is optimized; and finally, based on the optimized consensus graph, realizing cancer subtype identification by applying a spectral clustering algorithm. According to the method, the self-reinforcement learning strategy is introduced, the interference of noise on sample relation capture is effectively relieved, and the clustering performance is remarkably improved.

The application relates to the technical field of artificial intelligence and bioinformatics, in particular to a multi-source spatial multi-omics data integration method based on hierarchical graph contrast learning, which comprises the following steps: preprocessing acquired multi-source spatial multi-omics data, and constructing a spatial adjacency graph based on spatial distance; using a graph attention autoencoder to encode modality-specific features of the preprocessed multi-omics data based on the spatial adjacency graph, so as to obtain latent feature representation and negative representation of each spatial node; integrating the encoded latent feature representation and negative representation through hierarchical graph contrast learning; decoding and reconstructing the integrated features based on a neural network, and outputting reconstructed features for downstream tasks to minimize reconstruction error. The method can improve the full-scene data integration capability and improve the multi-omics data fusion precision.

The invention provides a larch pine caterpillar resistance gene mining method based on GWAS and multi-omics data, and belongs to the technical field of gene mining. The method comprises the following steps: S1, carrying out insect-resistant phenotype identification on larch individuals, and carrying out variation identification on the larch individuals based on a liquid-phase gene chip sequencing technology; s2, using a mixed linear model to perform correlation analysis by taking a plot position and phenotype observation time as an interaction environment, taking population density at different observation time as a covariable and taking a larch anti-deciduous pine caterpillar index as a phenotype, and identifying a candidate QTL region; and S3, carrying out transcriptome sequencing on resistance extreme individuals, carrying out variation recognition, comparing read segments to the candidate QTL region, and manually analyzing and annotating a gene structure in the QTL region by using IGV. According to the method disclosed by the invention, low-cost variation recognition, environment noise weakening whole genome association analysis and gene structure accurate annotation on complex genome species are realized.

The application relates to a biological age prediction method and device, an electronic device and a storage medium, wherein the biological age prediction method comprises the following steps: acquiring target multi-omics data of a to-be-detected individual; inputting each omics data in the target multi-omics data into a corresponding trained aging clock model for prediction to obtain each predicted biological age corresponding to each omics data; and combining the predicted biological ages to obtain a target biological age of the to-be-detected individual. According to the application, the accuracy of biological age evaluation is improved.

The invention relates to a CMAF model construction method and system fusing multi-modal data for survival analysis, and the method comprises the steps: obtaining data, and carrying out the preprocessing of the data, so as to obtain preprocessed data; constructing a CMAF model for survival analysis, wherein the CMAF model comprises a pathological image encoder, a multi-omics encoder, a modal confrontation module, a common attention module, a pyramid position code generator and a Kronecker product fusion layer; converting the preprocessed pathological image into embedded representation through a pathological image encoder to obtain pathological image embedding; adopting a self-normalized neural network as a multi-omics encoder, and converting the preprocessed multi-omics data into embedding representation to obtain multi-omics embedding; and performing modal alignment on the pathological image embedded data and the multi-omics embedded data to obtain pathological embedding and multi-omics embedding after modal alignment. Through the cross-modal attention fusion module, multi-modal data are effectively integrated, personalized treatment is supported, and intelligent development of medical treatment is promoted.

Provided are methods, systems and computer program product embodiments for analyzing multi-omic data using a tensor regression model for genome-wide association studies in the life sciences. The unique structure of tensor covariates is leveraged to find associations between the omics data and complex diseases. Within this framework, the excessive dimensionality is reduced to a manageable level, leading to efficient estimations and predictions. The method is superior to using classical regression techniques in genome-wide association studies, which are challenged by analyzing multi-dimensional and uniquely structured data from the health and life sciences, in which covariates can take on more intricate forms such as multi-dimensional arrays. Embodiments have multiple uses in genomics, proteomics, metabolomics, multi-omics data integration, drug discovery, personalized medicine and predictive modeling, demonstrating the versatility and importance of tensor regression models to understand the associations between omics data and complex diseases.

The invention belongs to the field of biotechnology and genomics, and particularly relates to a method and system for determining the chromosome cardinal number of macrobrachium rosenbergii based on multi-omics conjoint analysis. According to the method, de novo assembly sequencing data and Hi-C sequencing data are obtained, after primary assembly is completed, chromosome-level candidate assemblies are generated under the condition that the number of chromosomes is not preset through Hi-C interaction signals, and whole-genome collinearity comparison is carried out on the chromosome-level candidate assemblies and no less than two disclosed reference genomes; and in combination with quality constraints such as collinearity continuity, Hi-C boundary features and BUSCO / LAI, carrying out comprehensive decision and iterative convergence on candidate chromosome boundaries, and finally outputting chromosome cardinal numbers and reviewable evidence chains. The embodiment of the invention shows that the method can identify and correct the number redundancy caused by the over-resolution of the reference genome, determines the cardinal number of macrobrachium rosenbergii as n = 57 (2n = 114), and improves the objectivity and reliability of cardinal number determination.

The invention relates to the technical field of alfalfa screening, discloses a high-quality and high-yield alfalfa hybrid combination screening method fusing multiple technologies, provides a multi-dimensional coding and double-elite evolution mechanism, constructs an improved non-dominated sorting genetic algorithm-II, and provides a high-quality and high-yield alfalfa hybrid combination screening method by coding a hybrid combination into a continuous real number vector. A chromosome expression mode decoupled from a problem structure is constructed, a double-elite mechanism of elite retention and offset mating is introduced, the retention and heredity ability of a high-quality solution is improved, offset crossover and jitter operation is used, premature convergence is avoided, efficient global search of a strain combination space is achieved, and a high-quality solution is obtained. Meanwhile, a multi-omics association model of genes, metabolites and phenotypes is provided, transcriptome, metabolome and molecular marker data are comprehensively utilized, a multi-dimensional character prediction model is constructed, early-stage accurate screening of filial generations is achieved through correlation analysis and marker verification, the error selection rate is reduced, and the breeding period is shortened.

The invention discloses a multi-omics plant phenotype prediction method fused with small RNA, and relates to the field of bioinformatics and agricultural biology, and the method comprises the following steps: obtaining and preprocessing genotype, transcript expression and small RNA expression data to obtain a standardized feature matrix; performing feature screening based on Pearson correlation on the standardized feature matrix to obtain corresponding genotype, transcript expression and small RNA expression feature subsets; and based on a pre-trained MirGP prediction model, through multi-branch feature extraction, multistage efficient channel attention fusion and regression prediction, outputting a plant phenotype prediction result. According to the method, multiple types of small RNA regulation and control features are introduced and a hierarchical deep learning fusion architecture is constructed, so that multiple omics feature dimensions are expanded, data redundancy is reduced, and effective integration and deep mining of cross-modal features are realized.

This invention belongs to the interdisciplinary field of bioscience and information technology, and relates to a method for assessing the intensity of gene transcriptional regulation based on multi-omics data. Addressing the challenges of systematically integrating multi-omics data and quantitatively characterizing the correspondence between regulatory elements and genes, as well as the intensity of regulation, in existing transcriptional regulation analyses, this invention integrates epigenomics, transcriptomics, and three-dimensional genomics data to construct a site-gene regulatory intensity model. This method extracts candidate cis-regulatory sites and calculates their regulatory activities. It then determines the site-gene regulatory weights by combining distance weighting and three-dimensional genomic contact information, thereby obtaining the regulatory intensity at the site-gene pair and gene levels. An iterative algorithm distinguishes between positive and negative regulatory effects, achieving correction and summarization of regulatory intensity. This invention can quantitatively assess the regulatory intensity of cis-regulatory elements on target genes and the overall regulatory effect on genes, and is suitable for studying gene transcriptional regulation mechanisms and analyzing gene expression changes under different treatment conditions.

The present application relates to a disease target discovery method and system based on a multi-agent architecture. The multi-agent architecture comprises a multi-omics agent, a structure agent and a central agent. The disease target discovery method comprises: obtaining a task description in natural language and multi-modal data of a user by the central agent; analyzing multi-omics data by the multi-omics agent to generate a first initial target recommendation report, and analyzing structure data by the structure agent to generate a second initial target recommendation report; performing consensus analysis on the first initial target recommendation report and the second initial target recommendation report by the central agent, identifying consensus targets and unique targets, and feeding back the unique targets to the multi-omics agent and the structure agent respectively; verifying the unique targets by the multi-omics agent and the structure agent respectively to generate a first supplementary target recommendation report and a second supplementary target recommendation report; and generating a final target recommendation report by the central agent.

The invention provides a micro-fluidic chip for exosome multi-omics conjoint analysis, and belongs to the technical field of micro-fluidic analysis and detection. The micro-fluidic chip is sequentially provided with a micro-fluidic structure layer, a supporting layer and a liquid drop collecting layer from top to bottom in a stacked mode. The micro-fluidic structure layer comprises a liquid drop generation unit, a protein detection chip unit and a liquid drop collection port which are communicated in sequence, and the liquid drop generation unit is provided with a sample injection hole; the liquid drop collecting layer is provided with a liquid drop collecting chamber, and the liquid drop collecting chamber corresponds to the liquid drop collecting opening and is used for collecting liquid drops after protein detection and carrying out nucleic acid detection; the supporting layer is detachably connected with the liquid drop collecting layer, and the supporting layer is provided with liquid drop through holes corresponding to the liquid drop collecting cavities. According to the present invention, the exosome nucleic acid signal of the microfluidic chip can be stably detected, such that the nucleic acid analysis can be performed while the protein detection is completed so as to achieve the dual-mode and multi-omics detection, the sensitivity is superior to the sensitivity of the traditional method, and the disease diagnosis and detection accuracy is significantly improved.

The invention provides a multi-omics fusion phenotype prediction method and system based on a potential space of an auto-encoder, and belongs to the technical field of biological information and machine learning, and the method comprises the steps: carrying out the preprocessing of multi-source omics data, such as genotype, metabolome and hyperspectrum, and splicing the data into a high-dimensional matrix; carrying out unsupervised training on the matrix by using an auto-encoder, carrying out nonlinear compression on the matrix to a uniform low-dimensional potential space, and generating fusion features; constructing supervised learning models such as XGBoost and the like to predict agronomic characters by taking the fusion features as input; and by calculating the genetic advantage and fusion gain of the target character, recommending an optimal omics use strategy according to a preset decision rule. According to the method, heterogeneous multi-omics data are effectively integrated through the auto-encoder, noise is filtered out, a cross-omics nonlinear relation is mined, and the prediction precision of a complex character is remarkably improved. Meanwhile, the provided quantitative decision framework provides a scientific basis for efficient configuration of omics resources in breeding practice, and predictive performance and cost control are both considered.