اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدGraph Representation Learning on Tissue-Specific Multi-Omics
344
0
0.0
(
0
)
تأليف
Amine Amor
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Combining different modalities of data from human tissues has been critical in advancing biomedical research and personalised medical care. In this study, we leverage a graph embedding model (i.e VGAE) to perform link prediction on tissue-specific Gene-Gene Interaction (GGI) networks. Through ablation experiments, we prove that the combination of multiple biological modalities (i.e multi-omics) leads to powerful embeddings and better link prediction performances. Our evaluation shows that the integration of gene methylation profiles and RNA-sequencing data significantly improves the link prediction performance. Overall, the combination of RNA-sequencing and gene methylation data leads to a link prediction accuracy of 71% on GGI networks. By harnessing graph representation learning on multi-omics data, our work brings novel insights to the current literature on multi-omics integration in bioinformatics.
قيم البحث
اقرأ أيضاً
We present two algorithms designed to learn a pattern of correspondence between two data sets in situations where it is desirable to match elements that exhibit an affine relationship. In the motivating case study, the challenge is to better understa
nd micro-RNA (miRNA) regulation in the striatum of Huntingtons disease (HD) model mice. The two data sets contain miRNA and messenger-RNA (mRNA) data, respectively, each data point consisting in a multi-dimensional profile. The biological hypothesis is that if a miRNA induces the degradation of a target mRNA or blocks its translation into proteins, or both, then the profile of the former should be similar to minus the profile of the latter (a particular form of affine relationship). The algorithms unfold in two stages. During the first stage, an optimal transport plan P and an optimal affine transformation are learned, using the Sinkhorn-Knopp algorithm and a mini-batch gradient descent. During the second stage, P is exploited to derive either several co-clusters or several sets of matched elements. A simulation study illustrates how the algorithms work and perform. A brief summary of the real data application in the motivating case-study further illustrates the applicability and interest of the algorithms.
Genomics, especially multi-omics, has made precision medicine feasible. The completion and publicly accessible multi-omics resource with clinical outcome, such as The Cancer Genome Atlas (TCGA) is a great test bed for developing computational methods
that integrate multi-omics data to predict patient cancer phenotypes. We have been utilizing TCGA multi-omics data to predict cancer patient survival, using a variety of approaches, including prior-biological knowledge (such as pathways), and more recently, deep-learning methods. Over time, we have developed methods such as Cox-nnet, DeepProg, and two-stage Cox-nnet, to address the challenges due to multi-omics and multi-modality. Despite the limited sample size (hundreds to thousands) in the training datasets as well as the heterogeneity nature of human populations, these methods have shown significance and robustness at predicting patient survival in independent population cohorts. In the following, we would describe in detail these methodologies, the modeling results, and important biological insights revealed by these methods.
Advances in high-throughput sequencing technologies and bioinformatics approaches over almost the last three decades have substantially increased our ability to explore microorganisms and their functions-including those that have yet to be cultivated
in pure isolation. Genome-resolved metagenomic approaches have enabled linking powerful functional predictions to specific taxonomical groups with increasing fidelity. Additionally, whole community gene expression surveys and metabolite profiling have permitted direct surveys of community-scale functions in specific environmental settings. These advances have allowed for a shift in microbiome science away from descriptive studies and towards mechanistic and predictive frameworks for designing and harnessing microbial communities for desired beneficial outcomes. Here, we review how modern genome-resolved metagenomic approaches have been applied to a variety of water engineering applications from lab-scale bioreactors to full-scale systems. We describe integrated omics analysis across engineered water systems and the foundations for pairing these insights with modeling approaches. Lastly, we summarize emerging omics-based technologies that we believe will be powerful tools for water engineering applications. Overall, we provide a framework for microbial ecologists specializing in water engineering to apply cutting-edge omics approaches to their research questions to achieve novel functional insights. Successful adoption of predictive frameworks in engineered water systems could enable more economically and environmentally sustainable bioprocesses as demand for water and energy resources increases.
Recognizing multiple labels of images is a practical and challenging task, and significant progress has been made by searching semantic-aware regions and modeling label dependency. However, current methods cannot locate the semantic regions accuratel
y due to the lack of part-level supervision or semantic guidance. Moreover, they cannot fully explore the mutual interactions among the semantic regions and do not explicitly model the label co-occurrence. To address these issues, we propose a Semantic-Specific Graph Representation Learning (SSGRL) framework that consists of two crucial modules: 1) a semantic decoupling module that incorporates category semantics to guide learning semantic-specific representations and 2) a semantic interaction module that correlates these representations with a graph built on the statistical label co-occurrence and explores their interactions via a graph propagation mechanism. Extensive experiments on public benchmarks show that our SSGRL framework outperforms current state-of-the-art methods by a sizable margin, e.g. with an mAP improvement of 2.5%, 2.6%, 6.7%, and 3.1% on the PASCAL VOC 2007 & 2012, Microsoft-COCO and Visual Genome benchmarks, respectively. Our codes and models are available at https://github.com/HCPLab-SYSU/SSGRL.
Multi-omics data, that is, datasets containing different types of high-dimensional molecular variables (often in addition to classical clinical variables), are increasingly generated for the investigation of various diseases. Nevertheless, questions
remain regarding the usefulness of multi-omics data for the prediction of disease outcomes such as survival time. It is also unclear which methods are most appropriate to derive such prediction models. We aim to give some answers to these questions by means of a large-scale benchmark study using real data. Different prediction methods from machine learning and statistics were applied on 18 multi-omics cancer datasets from the database The Cancer Genome Atlas, containing from 35 to 1,000 observations and from 60,000 to 100,000 variables. The considered outcome was the (censored) survival time. Twelve methods based on boosting, penalized regression and random forest were compared, comprising both methods that do and that do not take the group structure of the omics variables into account. The Kaplan-Meier estimate and a Cox model using only clinical variables were used as reference methods. The methods were compared using several repetitions of 5-fold cross-validation. Unos C-index and the integrated Brier-score served as performance metrics. The results show that, although multi-omics data can improve the prediction performance, this is not generally the case. Only the method block forest slightly outperformed the Cox model on average over all datasets. Taking into account the multi-omics structure improves the predictive performance and protects variables in low-dimensional groups - especially clinical variables - from not being included in the model. All analyses are reproducible using freely available R code.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
