Historical Note
Lessons Learned as President of the Institute for Systems Biology (2000–2018)Open access
Leroy E. Hood
View
abstract
系统生物学及P4医学创始人,美国国家科学奖章获得者,美国科学院、工程科学院及医学科学院三院院士Leroy Hood博士回顾创立世界第一家系统生物学研究院(ISB)并担任院长17年来的经验教训。讲述了ISB通过技术创新以及系统性思维及手段解析复杂生物体系,以及在大数据背景下催生的全新系统P4医学理念。阐述了ISB在十几年时间内发展为世界级科研机构的跨学科背景、合作氛围、领导及组织结构、发展的关键结点、以及建立战略合作伙伴的重要性。提出21世纪医学将由关注疾病转向关注健康的P4医疗体系。2016年ISB加入美国最大的非盈利医疗体系之一Providence St. Joseph Health,使P4医学的临床践行成为可能。此文对在建的系统生物医学、精准医学、及转化医学机构的建设,具有很强的借鉴意义。
Page 1-9
Review Article
Genome Writing: Current Progress and Related Applications
Yueqiang Wang, Yue Shen, Ying Gu, Shida Zhu, Ye Yin
View
abstract
The ultimate goal of synthetic biology is to build customized cells or organisms to meet specific industrial or medical needs. The most important part of the customized cell is a synthetic genome. Advanced genomic writing technologies are required to build such an artificial genome. Recently, the partially-completed synthetic yeast genome project represents a milestone in this field. In this mini review, we briefly introduce the techniques for de novo genome synthesis and genome editing. Furthermore, we summarize recent research progresses and highlight several applications in the synthetic genome field. Finally, we discuss current challenges and future prospects.
生物学研究的终极目标是改造生物体和自然环境,使之符合人类的利益。充分认识生物体生长发育的分子机制,是实施人为改造的前提和基础。在长达数百年的历史中,生物学研究形成多个不同的学科,分别在不同维度上研究生物生长发育以及生态系统运行的机制。20世纪50年代,DNA双螺旋结构被发现,之后生命科学研究进入到分子生物学时代。随着时间推移,对小鼠、果蝇、拟南芥、酵母、大肠杆菌等少数几种模式生物的生长发育的分子机制的研究越来越透彻。在此基础上,人类逐渐能够做到根据自己意图来改造生物体,甚至创造新生物体。合成生物学,作为一门新学科因此而诞生。传统生物学研究,注重对自然界内在机制的“理解”;合成生物学则注重对自然界的“改造”。由于生物体的性状主要由其基因组DNA控制,因而“改造”的过程主要体现为对生物体基因组DNA的改写。
在本综述中,我们以基因组写(Genome writing)技术为核心,主要介绍用于改写生物基因组的相关技术,同时也介绍了合成生物学未来可能的重要应用。关于基因组写技术,我们介绍了如何在大尺度和小尺度上改写基因组。大尺度基因组写技术,主要介绍了如何从引物阶段,通过逐级组装从而获得一个合成的基因组。小尺度基因组写技术,主要介绍了利用基因编辑技术对基因组进行定点精确编辑。在应用研究方面,我们介绍了利用合成生物学技术手段构建人源化动物、研究人类基因组突变体功能,等方面的案例。合成生物学的发展也带来一些新的伦理和安全的问题,在文章的展望部分我们对相关问题作了简要讨论。
Page 10-16
Review Article
Deep Learning and Its Applications in Biomedicine
Chensi Cao, Feng Liu, Hai Tan, Deshou Song, Wenjie Shu, Weizhong Li, Yiming Zhou, Xiaochen Bo, Zhi Xie
View
abstract
Advances in biological and medical technologies have been providing us explosive volumes of biological and physiological data, such as medical images, electroencephalography, genomic and protein sequences. Learning from these data facilitates the understanding of human health and disease. Developed from artificial neural networks, deep learning-based algorithms show great promise in extracting features and learning patterns from complex data. The aim of this paper is to provide an overview of deep learning techniques and some of the state-of-the-art applications in the biomedical field. We first introduce the development of artificial neural network and deep learning. We then describe two main components of deep learning, i.e., deep learning architectures and model optimization. Subsequently, some examples are demonstrated for deep learning applications, including medical image classification, genomic sequence analysis, as well as protein structure classification and prediction. Finally, we offer our perspectives for the future directions in the field of deep learning.
近年来,高通量技术的发展为我们提供了海量的生物学、医学、生理学数据,给生物医学发展带来了前所未有的机遇。同时,如何从这些体量大、异质性高、产生速度快的大数据中提取出能指导生物医学研究、辅助临床决策的信息,是数据分析面临的巨大挑战。人工智能方法已被广泛应用于生物医学大数据的处理,而为人工智能领域带来重大突破的深度学习,更是为生物医学大数据的分析提供了广阔前景。
1958 年,David Hubel 和Torsten Wiesel 在 Johns Hopkins University,研究瞳孔区域与大脑皮层神经元的对应关系。他们在猫的后脑头骨上开了一个3 毫米的小洞,向洞里插入电极,用于观察小猫看到各种形状、亮度、姿态的物体时大脑神经元的活跃程度。实验证明,猫由瞳孔感受到刺激,刺激信息紧接着被分层传输到后脑皮层中相应区域的视觉神经元。受到此过程的启发,人工神经网络(ANN)通过分层的人造神经元来模拟大脑对物体的感知。
最初,由于结构浅且计算能力有限,ANN在实际应用中曾一度落后于其他简单的机器学习算法,如支持向量机(SVM),随机森林和k-最近邻算法等。近年来,由于计算机能力和方法的改进,图形处理单元(GPU)和反向传播算法(BP)的使用又重新促进了ANN的发展。并且人们发现,更多分层的ANN可以提取更复杂的特征,提高对事物感知的准确性从而发展为深度学习算法。
深度学习是机器学习领域的一个新兴和快速发展领域。它试图通过多层深度神经网络(DNN)对大规模数据进行抽象建模,从而理解图像,声音和文本等数据。一般而言,深度学习有两个属性:(1)多层非线性处理单元,以及(2)有监督或无监督的特征学习。深度学习算法(DL)是一类算法的统称。虽然这些算法基本假设和理论不同,但特征提取的基本思想和过程非常相似。在正向传播过程中,网络接收到输入信息,然后沿着网络连接将输入信息传输到最终层产生预测或重建结果。在反向传递过程中,通过最小化预测数据与实际数据之间的差异来优化网络的连接参数。在这篇文章中,作者介绍了深度神经网络中的分层及其理论依据,如何形成深度神经网络,信息如何传递到最终层,如何定义预测数据与实际数据之间的差异,以及如何利用差异来优化网络的连接参数。通过对这些算法的了解,读者可以根据自己的数据特点,选择合适的深度学习模型,调整算法结构,以期提高模型的准确性和泛化性。
由于对信息处理的高效、准确,深度学习算法近年来已广泛应用于自动语音识别,图像识别,自然语言处理等领域。与此同时,由于高通量技术的发展,生物医学数据大幅增长,亟需有效且高效的计算工具来存储,分析和解释。深度学习的出现和应用正好为此提供了良好契机。在医学图像识别领域中,深度学习已在皮肤癌、致盲性眼病、乳腺癌、糖尿病视网膜病和肺炎的诊断中大显身手,其表现匹敌各领域专家。对于爆炸式增长的组学数据,深度学习亦游刃有余:预测基因表达量,预测蛋白质结合位点、可变剪切位点,解读蛋白质三维结构,甚至也为非编码区的解读提供了可能。在上述这些领域的应用中,深度学习的表现都远远超过了其他人工智能算法,准确率的提升高达50%。在这篇文章中,作者总结了深度学习的多种模型在不同的生物医学领域中的应用。
除了目前所取得的丰硕成果,深度学习在生物医学领域中的应用也面临着诸多挑战:获取高质量的、大量的样本标注需要巨大的人力物力;医学数据往往具有不平衡性、异质性;对于医学诊断类的应用场景,模型需要具有更高的敏感性。种种挑战为深度学习模型提出了更高的要求。在未来,我们期望深度学习在生物医学领域中取得更深远的影响,不断促进精准医疗领域的发展。
Page 17-32
Review Article
Human Gut Microbiota and Gastrointestinal Cancer
Changting Meng, Chunmei Bai, Thomas D. Brown, Leroy E. Hood, Qiang Tian
View
abstract
Human gut microbiota play an essential role in both healthy and diseased states of humans. In the past decade, the interactions between microorganisms and tumors have attracted much attention in the efforts to understand various features of the complex microbial communities, as well as the possible mechanisms through which the microbiota are involved in cancer prevention, carcinogenesis, and anti-cancer therapy. A large number of studies have indicated that microbial dysbiosis contributes to cancer susceptibility via multiple pathways. Further studies have suggested that the microbiota and their associated metabolites are not only closely related to carcinogenesis by inducing inflammation and immune dysregulation, which lead to genetic instability, but also interfere with the pharmacodynamics of anticancer agents. In this article, we mainly reviewed the influence of gut microbiota on cancers in the gastrointestinal (GI) tract (including esophageal, gastric, colorectal, liver, and pancreatic cancers) and the regulation of microbiota by diet, prebiotics, probiotics, synbiotics, antibiotics, or the Traditional Chinese Medicine. We also proposed some new strategies in the prevention and treatment of GI cancers that could be explored in the future. We hope that this review could provide a comprehensive overview of the studies on the interactions between the gut microbiota and GI cancers, which are likely to yield translational opportunities to reduce cancer morbidity and mortality by improving prevention, diagnosis, and treatment.
肠道微生物与人体健康和多种疾病密切相关,这是近年来全世界科学家们研究的焦点之一。各种高通量测序和组学研究产生的大数据使得研究者们日益深入、广泛、精准地了解人体微生物与不同疾病的相关性及其机制。本文着重回顾总结了肠道微生物与胃肠道肿瘤的致癌和抑癌相关性基础研究和临床转化研究。该领域大量研究证实,人体的肠道微生物生态环境错综复杂,一方面,肠道微生物失调参与了慢性炎症反应、免疫失调、基因的不稳定性形成等过程,对于增强肿瘤易感性起到重要作用;另一方面,调控肠道微生物在肿瘤预防及治疗中可以发挥其独特的积极作用。本文所涉及的胃肠道肿瘤包括食管癌、胃癌、大肠癌、肝癌、胆管癌及胰腺癌等,所采用的文献内容涵括了肠道微生物在健康人群、癌前病变、各种胃肠道肿瘤中的特点;肠道微生物参与肿瘤性疾病发生发展的可能机制的探讨,参与肿瘤预防、肿瘤化疗、靶向治疗、免疫治疗等方面的探索性基础研究和临床观察性研究;并归纳总结了应用食物、肠道益生菌、益生元、抗生素以及中药等调控肠道菌群等相关研究。人体微生物与人类息息相关,监测人体微生物组,预测人体健康和疾病状态,调控肠道微生物推动关于肿瘤预防、诊断和治疗领域的新策略形成,从而实现以“预见性、预防性、个体化及参与性医学模式”之理念指导和推动微生物研究,进一步降低肿瘤发病率和致死率。
Page 33-49
Review Article
Microvesicles as Emerging Biomarkers and Therapeutic Targets in Cardiometabolic Diseases
Yan Chen, Guangping Li, Ming-Lin Liu
View
abstract
Microvesicles (MVs, also known as microparticles) are small vesicles that originate from plasma membrane of almost all eukaryotic cells during apoptosis or activation. MVs can serve as extracellular vehicles to transport bioactive molecules from their parental cells to recipient target cells, thereby serving as novel mediators for intercellular communication. Importantly, more and more evidence indicates that MVs could play important roles in early pathogenesis and subsequent progression of cardiovascular and metabolic diseases. Elevated plasma concentrations of MVs, originating from red blood cells, leukocytes, platelets, or other organs and tissues, have been reported in various cardiometabolic diseases. Circulating MVs could serve as potential biomarkers for disease diagnosis or therapeutic monitoring. In this review, we summarized recently-published studies in the field and discussed the role of MVs in the pathogenesis of cardiometabolic diseases. The emerging values of MVs that serve as biomarker for non-invasive diagnosis and prognosis, as well as their roles as novel therapeutic targets in cardiometabolic diseases, were also described.
细胞微囊泡(microvesicles, MVs) 在1967年首先被wolf报道,认为它是血小板释放的细胞膜囊泡,并称之为“血小板尘埃”。随后研究又陆续发现,除血小板外,循环血液中的红细胞、中性粒细胞、淋巴细胞以及单核细胞,血管内皮细胞,部分平滑肌细胞在不同的刺激条件下均能释放此类物质。目前认为MVs是多种细胞来源的,在细胞激活或凋亡时产生的含有细胞膜结构的物质。它可以通过介导配体受体反应或传递胞浆成分及细胞器等方式,参与并影响靶细胞内信号转导通路,从而介导母细胞和靶细胞的联系。MVs的释放是机体的一种自我调节机制,可以介导细胞之间的信号传导和免疫调节,在肿瘤、心血管疾病、神经系统疾病病理生理学和发病机制中起重要的作用。MVs释放到血液细胞中, 作为一种新兴的标记物和调节介质,并且随着对其的研究深入,关于MVs作为疾病的生物标记物和液体活体检测靶点的期望越来越高。因此,本文就MVs的生物学特性,细胞间信息传递方式,以及在心血管代谢疾病中作为生物标记物或者治疗靶点的研究进展进行综述。
Page 50-62
Review Article
Preprocessing of 2-Dimensional Gel Electrophoresis Images Applied to Proteomic Analysis: A Review
Manuel Mauricio Goez, Maria Constanza Torres-Madroñero, Sarah Röthlisberger, Edilson Delgado-Trejos
View
abstract
Various methods and specialized software programs are available for processing two-dimensional gel electrophoresis (2-DGE) images. However, due to the anomalies present in these images, a reliable, automated, and highly reproducible system for 2-DGE image analysis has still not been achieved. The most common anomalies found in 2-DGE images include vertical and horizontal streaking, fuzzy spots, and background noise, which greatly complicate computational analysis. In this paper, we review the preprocessing techniques applied to 2-DGE images for noise reduction, intensity normalization, and background correction. We also present a quantitative comparison of non-linear filtering techniques applied to synthetic gel images, through analyzing the performance of the filters under specific conditions. Synthetic proteins were modeled into a two-dimensional Gaussian distribution with adjustable parameters for changing the size, intensity, and degradation. Three types of noise were added to the images: Gaussian, Rayleigh, and exponential, with signal-to-noise ratios (SNRs) ranging 8–20 decibels (dB). We compared the performance of wavelet, contourlet, total variation (TV), and wavelet-total variation (WTTV) techniques using parameters SNR and spot efficiency. In terms of spot efficiency, contourlet and TV were more sensitive to noise than wavelet and WTTV. Wavelet worked the best for images with SNR ranging 10–20 dB, whereas WTTV performed better with high noise levels. Wavelet also presented the best performance with any level of Gaussian noise and low levels (20–14 dB) of Rayleigh and exponential noise in terms of SNR. Finally, the performance of the non-linear filtering techniques was evaluated using a real 2-DGE image with previously identified proteins marked. Wavelet achieved the best detection rate for the real image.
蛋白质组学是分析在生物体内产生的一些列蛋白质,包括分析蛋白表达、蛋白存在与否及其丰度的直接测量等,由此来整体而全面理解细胞进程和识别药物靶点、诊断和预后标志物。在比较蛋白质组学研究中应用最广的技术就是二维凝胶电泳(2-DGE)图像技术。目前已经有各种各样的方法和专门的软件程序来处理2-DGE图像,如MELANIE和PDQuest等。然而,由于2-DGE图像通常存在一些异常现象,一个可靠的、自动化、和高度可再现的2-DGE图像分析系统仍未完成。在2-DGE图像最常见的异常现象包括垂直和水平图像拖尾、模糊点以及背景噪音,这些异常大大增加了计算分析的复杂性。在本文中,我们回顾了2-DGE图像的预处理技术,主要包括2-DGE图像的降噪、密度标准化和背景校正。通过分析在特定条件下滤波器的性能,我们定量地比较了应用于合成凝胶图像的非线性滤波技术。合成蛋白质被建模成一个二维高斯分布,并且可通过调整参数改变蛋白图像的大小、密度和退化。高斯、瑞利和指数这三种类型的噪声被添加到图片,其信噪比范围在8-20分贝(dB)。我们通过使用参数信噪比和点效率来比较了小波、轮廓波、总变异以及小波总变异技术的性能。就点效率而言,轮廓波和总变异比小波和小波总变异对噪声更敏感。小波能最好地处理信噪比范围在10 - 20分贝的图像,而小波总变异则能更好地处理高噪音水平的图像。就信噪比而言,小波也呈现出最佳性能来处理任何程度的高斯噪声、低水平(20-14 dB)的瑞利噪声和指数噪声。最后, 通过使用已经确定蛋白质标记的真实2-DGE图片来评估非线性滤波技术的性能,我们发现小波技术能够取得真实图片的最好检出率。
Page 63-72
Original Research
The Immunome of Colon Cancer: Functional In Silico Analysis of Antigenic Proteins Deduced from IgG Microarray Profiling
Johana A. Luna Coronell, Khulan Sergelen, Philipp Hofer, István Gyurján, Stefanie Brezina, Peter Hettegger, Gernot Leeb, Karl Mach, Andrea Gsur, Andreas Weinhäusel
View
abstract
Characterization of the colon cancer immunome and its autoantibody signature from differentially-reactive antigens (DIRAGs) could provide insights into aberrant cellular mechanisms or enriched networks associated with diseases. The purpose of this study was to characterize the antibody profile of plasma samples from 32 colorectal cancer (CRC) patients and 32 controls using proteins isolated from 15,417 human cDNA expression clones on microarrays. 671 unique DIRAGs were identified and 632 were more highly reactive in CRC samples. Bioinformatics analyses reveal that compared to control samples, the immunoproteomic IgG profiling of CRC samples is mainly associated with cell death, survival, and proliferation pathways, especially proteins involved in EIF2 and mTOR signaling. Ribosomal proteins (e.g., RPL7, RPL22, and RPL27A) and CRC-related genes such as APC, AXIN1, E2F4, MSH2, PMS2, and TP53 were highly enriched. In addition, differential pathways were observed between the CRC and control samples. Furthermore, 103 DIRAGs were reported in the SEREX antigen database, demonstrating our ability to identify known and new reactive antigens. We also found an overlap of 7 antigens with 48 “CRC genes.” These data indicate that immunomics profiling on protein microarrays is able to reveal the complexity of immune responses in cancerous diseases and faithfully reflects the underlying pathology.
根据流行病学研究,结直肠癌癌发生率在女性肿瘤排第二和男性肿瘤排第三,是一种世界排名靠前的高发性恶性肿瘤。目前结直肠癌临床检测有多种方法,如直肠镜检测、乙状结直肠镜检测等,这种有创检测不仅带给患者心理上的压力而且造成身体伤害。识别敏感而特异markers对于结直肠癌的无创检测迫在眉睫。已知在肿瘤转化以及进展过程中蛋白表达改变可以造成免疫反应和诱导肿瘤自身抗原形成。免疫反应可以通过免疫编辑来促进肿瘤增殖,而免疫编辑影响了循环抗体的成分和数量。抗体是一种可被用于临床癌症早诊检测且非常稳定markers。因此,研究结肠癌免疫组特征及其来源不同反应的抗原(DIRAGs)的自身抗体特征可能对该疾病相关的异常细胞机制或富集的生物学功能网络和临床诊断提供新的视角。为了描述结直肠癌的抗体谱特征,我们先从15417个健康人包含6369独特人蛋白的cDNA表达克隆的蛋白质微阵列来识别DIRAGs。然后用32结直肠癌(CRC)患者的血浆样本的IgG和32个对照的IgG的蛋白质微阵列数据进行比较分析。我们识别了671个特异的DIRAGs和632个在CRC中更高活性的DIRAGs。进一步分析表明,不同生物学功能通路也被发现在CRC组和对照组之间。相比于对照组,CRC的IgG组谱主要与细胞凋亡、生存以及增值通路相关,尤其是蛋白质参与EIF2和mTOR信号通路。这些DIRAGs高度富集在核糖体蛋白(如RPL7,RPL22,RPL27A)和CRC-related基因(如APC,AXIN1,E2F4,MSH2 PMS2,TP53)。同时,我们利用公共数据库分析发现103个 DIRAGs已经被报道在SEREX抗原数据库,这表明我们可以同时识别已知抗原并发现新活性抗原。这些数据表明基于蛋白质微阵列得到的免疫组谱能够揭示癌症疾病免疫反应的复杂性以及如实地反映潜在的病理。
La caracterización del immunome del cáncer de colon y su perfil de anticuerpos de antígenos con reacción diferencial (DIRAGs, en inglés) podría proporcionar información sobre mecanismos celulares aberrantes o redes enriquecidas asociadas con enfermedades. El propósito de este estudio fue caracterizar el perfil de anticuerpos de muestras de plasma de 32 pacientes con cáncer de colon (CC) y 32 controles utilizando proteínas extraídas de 15,417 clones de expresión de ADNc humano en microarrays. Se identificaron 671 DIRAG únicos y 632 resultaron más altamente reactivos en muestras de CC. Análisis bioinformáticos revelan que en comparación con las muestras de control, el perfil inmunoproteómico de IgG de las muestras de CC se asocia principalmente con la muerte celular, la supervivencia y las vías de proliferación, especialmente las proteínas implicadas en la señalización de EIF2 y mTOR. Las proteínas ribosómicas (por ejemplo, RPL7, RPL22 y RPL27A) y los genes relacionados con CC tales como APC, AXIN1, E2F4, MSH2, PMS2 y TP53 se encontraron altamente enriquecidos. También se observaron diferenciales en las vías de señalización entre las muestras de CC y las de control. Además, se encontraron 103 DIRAGs reportados en la base de datos de antígenos de SEREX, lo que demuestra nuestra capacidad para identificar antígenos reactivos conocidos y nuevos. También encontramos una superposición de 7 antígenos con 48 "genes de CC". Estos resultados indican que el perfilamiento immunómico en microarrays de proteínas es capaz de revelar la complejidad de la respuesta inmune en enfermedades cancerosas y refleja fielmente la patología subyacente.
Page 73-84