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.
Integrate Any Omics: Towards genome-wide data integration for patient stratification
Peter Munk Cardiac Centre, University Health Network(彼得·蒙克心脏中心,大学健康网络);Vector Institute for Artificial Intelligence(向量人工智能研究所);Department of Computer Science, University of Toronto(多伦多大学计算机科学系);Department of Molecular Genetics, University of Toronto(多伦多大学分子遗传学系);Princess Margaret Cancer Centre, University Health Network(玛格丽特公主癌症中心,大学健康网络);Department of Medical Biophysics, University of Toronto(多伦多大学医学生物物理学系);Genetics and Genome Biology, the Hospital for Sick Children(病童医院遗传学与基因组生物学)
CMGL: Confidence-guided Multi-omics Graph Learning for Cancer Subtype Classification
College of Computer Science, Sichuan University(四川大学计算机学院);Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences(中国科学院新疆物理化学研究所);College of Mathematics, Sichuan University(四川大学数学学院);School of Artificial Intelligence and Data Science, University of International Business and Economics(国际经贸大学人工智能与数据科学学院)
Optimizing Prognostic Biomarker Discovery in Pancreatic Cancer Through Hybrid Ensemble Feature Selection and Multi-Omics Data
Department of Cancer Genetics, Institute for Cancer Research, Oslo University Hospital, Oslo, Norway(癌症遗传学系,癌症研究所,奥斯陆大学医院,奥斯陆,挪威);Oslo Centre for Biostatistics and Epidemiology (OCBE), Department of Biostatistics, University of Oslo, Oslo, Norway(奥斯陆生物统计与流行病学中心(OCBE),生物统计学系,奥斯陆大学,奥斯陆,挪威);Department of Informatics, University of Oslo, Oslo, Norway(信息学系,奥斯陆大学,奥斯陆,挪威);Department of Statistics, Ludwig Maximilian University of Munich, Munich, Germany(统计学系,慕尼黑路德维希-马克西米利安大学,慕尼黑,德国);Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心(MCML))
Integrative multi-omics module network inference with Lemon-Tree
模块网络推断是一种已建立的统计方法,用于从集成的多组学数据集重建共表达模块及其上游调控程序,所述集成的多组学数据集测量跨动态过程的不同个体、实验条件或时间点的各种细胞组分的活性水平。我们已经开发了Lemon-Tree,这是一个开源的、平台无关的、模块化的、可扩展的软件包,它实现了用于模块网络推理的最先进的集成方法。我们使用大规模肿瘤数据集对Lemon-Tree进行了基准测试,并表明Lemon-Tree算法与最先进的模块网络推理软件相比毫不逊色。我们还分析了来自The Cancer Genome Atlas的胶质母细胞瘤样本中测量的体细胞拷贝数改变和基因表达水平的大型数据集,发现Lemon-Tree正确识别了已知的胶质母细胞瘤癌基因和肿瘤抑制因子作为推断模块网络中的主调节因子。Lemon-Tree预测的新候选驱动基因使用肿瘤通路和存活分析进行验证。Lemon-Tree可以在GNU通用公共许可证2.0版下从http://lemon-tree.googlecode.com获得。
UnPaSt: unsupervised patient stratification by biclustering of omics data
University of Hamburg(汉堡大学);Technical University of Munich(慕尼黑工业大学);Netherlands Cancer Institute(荷兰癌症研究所);ACMetric;University of Campinas(坎皮纳斯大学);University of British Columbia(不列颠哥伦比亚大学);Friedrich-Alexander-Universität Erlangen-Nürnberg(弗里德里希-亚历山大大学埃尔兰根-纽伦堡);Vrije Universiteit Amsterdam(阿姆斯特丹自由大学);Altius Institute for Biomedical Sciences(Altius生物医学科学研究所);University of Liège(列日大学);Aarhus University(奥胡斯大学);Aarhus University Hospital(奥胡斯大学医院);University Medical Center Hamburg-Eppendorf(汉堡-埃彭多夫大学医学中心);Simon Fraser University(西蒙弗雷泽大学);Vancouver Prostate Centre(温哥华前列腺中心);Leibniz Institute of Virology(莱布尼茨病毒学研究所);University of Southern Denmark(南丹麦大学)
Prediction approaches for partly missing multi-omics covariate data: A literature review and an empirical comparison study
University of Munich(慕尼黑大学);Max Planck Institute of Psychiatry(马克斯·普朗克精神病学研究所);International Max Planck Research School for Translational Psychiatry(国际马克斯·普朗克转化精神病学研究学院)
Heterogeneous Graph Attention Network With Multimodal Feature Selection for Cancer Multiomics Integration
School of Computer Science, The University of Sheffield(谢菲尔德大学计算机学院);Centre for Machine Intelligence, The University of Sheffield(谢菲尔德大学机器智能中心);School of Biosciences, The University of Sheffield(谢菲尔德大学生物科学学院)
MLOmics: Cancer Multi-Omics Database for Machine Learning
Bioinformatics Center, Institute for Chemical Research, Kyoto University(京都大学生物信息学中心、化学研究所);SANKEN, Osaka University(大阪大学SANKEN);IRCN, The University of Tokyo(东京大学IRCN);Institute for Quantitative Biosciences, The University of Tokyo(东京大学定量生物科学研究所);Department of Computer Science, University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校计算机科学系);Carle Illinois College of Medicine, University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校Carle医学院)
Deep radiomic signature with immune cell markers predicts the survival of glioma patients
Lady Davis Institute for Medical Research, McGill University(麦吉尔大学戴维斯夫人医学研究所);Hudson Institute of Medical Research(哈德逊医学研究所);Montreal Neurological Institute, McGill University(麦吉尔大学蒙特利尔神经学研究所);Perelman School of Medicine, University of Pennsylvania(宾夕法尼亚大学佩雷尔曼医学院);The Laboratory for Imagery, Vision and Artificial Intelligence(图像、视觉与人工智能实验室)