1/U19A200225

安徽酒精性脂肪肝病发病机制与预防干预靶标

Mechanism of onset and progression and targets of prevention and intervention of alcoholic liver disease in Anhui province

酒精性脂肪肝

危险因素

生物标志物

表观遗传

alcoholic liver disease

risk factors

biomarkers

epigenetics

中文摘要：（限400字）

安徽地区高发酒精性脂肪肝病（ALD），但是其临床特点和作用机制还不很清楚。前期研究表明在细胞动物水平和东亚人群中酒精代谢物、细胞色素P450s、活性氧自由基（ROS）、细胞因子、炎性因子、肠道菌群、肝炎病毒、酒精代谢网络有关酶及其基因与ALD发生发展机制及预防干预相关。为此，本项目首先聚焦安徽16个地级市人群，分析ALD支持/危险因素并调查生物标志物；接着在动物和细胞水平认识影响ALD发生发展和防治中脂质生成、氧化应激和炎症的遗传和表观遗传机制，筛选并验证支持/危险因素和生物标志物；最终回到人群，通过前瞻性观察验证支持因素和生物标志物。把影像学（USG、CAP、MRI、PET）、生物组学（代谢组，宏基因组，RNA-seq，蛋白质组）、核酸-蛋白互作分析（EMSA、ChIP-chip）、[表观]遗传（SNP、GWAS、去乙酰化）相结合。项目的完成有助于认识和防治安徽地区ALD，提高健康水平。

英文摘要：（限4000字符）

Alcoholic fatty liver disease (ALD) is prevalent in Anhui province, but its clinical characteristics and mechanism are still unclear. Previous studies have shown that alcoholic metabolites, cytochrome P450s, reactive oxygen species (ROS), cytokines, inflammatory factors, intestinal flora, hepatitis viruses, enzymes related to alcohol metabolism network and their genes are related to the occurrence, development, prevention and intervention of ALD. Therefore, the project first focused on the population of 16 prefecture-level cities in Anhui Province, analyzed ALD support/risk factors and investigated biomarkers; then assayed the genetic and epigenetic mechanisms affecting oxidative stress and inflammation in the occurrence and development of ALD, at the animal and cell levels, screened and validated support/risk factors and biomarkers; finally returned to the population, validated supporting factors and biomarkers through prospective observation. Imaging (USG, CAP, MRI, PET), bio-omics (metabolome, metagenome, RNA-seq, proteome), nucleic acid-protein interaction analysis (EMSA, ChIP-chip), [epi]genetics (SNP, GWAS, [de]acetylation) were combined. The completion of the project will help to understand and prevent ALD in Anhui and improve health outcomes.

2/9194230015

肝肠“免疫共振图谱”与炎症调控

Enterohepatic "immune resonance spectra" and inflammation modulating

“代谢器官”（肝、肾、肠、脑等）与免疫器官（胸腺、骨髓等）在免疫学特性上存在较大区别。“代谢器官”含有特定的细胞亚群和功能分子，形成区域免疫特性。区域免疫特性与所在“代谢器官”相关的众多疾病的发生发展紧密相关，由于对“代谢器官”区域免疫特性与疾病关系研究较少，影响了疾病免疫防治的发展。为了阐释炎症的区域免疫机制，申请人提出免疫共振图谱，研究具有区域免疫特性的“代谢器官”与免疫器官发生“共振”产生的模式及迁移；代谢、免疫器官，分别充当原子、静磁场；区域免疫与专业免疫间，类似核磁共振；代谢与专业免疫间，类似电子顺磁共振；免疫、细胞因子的演化、流动，类似磁共振成像、代谢流。项目整合多种技术（微生物组学、脂质组学、表观遗传学、等离子体共振、单细胞多功能强度指数等）深入解析肠-肝区域免疫特性、交互作用和关联研究，揭示恒定自然杀伤T细胞与炎症的联系，提出幼年时期营养-免疫干预减少成年肥胖与炎症的策略。

英文摘要：（限4000字符）

The immune traits of "metabolic organs" (liver, kidney, gut, brain, etc.) are different from those of immune organs (thymus, bone marrow, etc.). Metabolic organs contain specific cell subgroups and functional molecules, displaying regional immunity. Regional immunity is related to the onset and progression of high-incidence diseases of metabolic organs; less studies on relationship between regional immunity and metabolic diseases block immune therapy.

In the era of -omics, the applicant defines the metabolic resonance spectrum and immune resonance spectrum. Metabolic resonance spectra focus on the metabolic pattern and comparative and/or dynamic responses of biomaterials to perturbations (Zhao X.J. 2011); Nuclear Magnetic Resonance spectra underline amino acids, organic acids, nucleotides, etc; Electron Paramagnetic Resonance spectra emphasize free radicals and transition metal ions; Magnetic Resonance Imaging spectra underline spatial distribution of metabolites. Thus, immune resonance spectra explore the immune pattern and comparative and/or dynamic responses between "metabolic organs" and immune organs (Zhao X.J. 2019); resonance of regional immunity and professional immunity, resembles NMR; resonance of metabolism (non-regional immunity) and professional immunity, resembles EPR; distribution and flux of immune and cell factors, resembles MRI and metabolic flux.

The current study integrates microbiomics, lipidomics, epigenitics, surface plasmon resonance, polyfunctional strength index of single cells, and studies enterohepatic regional immunity, interaction and association, to explore the effect of invariant natural killer T cells on inflammation, and to provide an important basis on the rational nutritional supplement and immunity in neonates and infants against adult obesity and inflammation.

炎症 免疫调控 脂质组学 个体早期发育 表观遗传调控

Inflammation  Immune modulation  Lipidomics  Individual early development  Epigenetic tuning

3/7194100202

基于大数据的中国人口空间分布态势预测

Prediction of Population Spatial Distributions and Trends Based on Big Data in China

人口  空间分布  时间序列  大数据

Population

Spatial Distributions

Time Series

Big Data

中文摘要：（限400字）

人口分布影响国民经济和社会发展的可持续性，是微观主体、中观行业和宏观管理的决定因素。中国人口分级统计，有中央、省级、地级、县级、乡级和村级六级行政区划，每级有不同的城乡和民族区域自治类型，而且有常住与户籍人口之分；目前的人口统计存在不完整、不一致、不标准和不精准的问题。为了保证数据的精准、完整和一致，我们采用全国和各省统计年鉴的数据对上一年全国、各省、各地的人口进行校正，同时纳入国家卫生健康、资源环境的数据。依此评估我国人口空间分布现状及特征；把经济社会、资源环境、医疗卫生等可塑性因素纳入视野，通过多种半参数和非参数的相关与回归方法寻找影响人口分布变化的主要因素。以主要影响因素的变化为输入，使用多种有监督的时间序列方法建模、验证，选出最优模型预测人口空间分布趋势。纵观人口空间分布与时间演进，在国家战略的背景和约束条件下，以可塑性因素为重点提出“十四五”时期中国人口和城乡区域格局的政策建议。

英文摘要：（限4000字符）

Population distributions confer the sustainability of national economic and social developments, and are important major determinants of micro-entities, middle-industries and macro-managements. China's population are counted by different levels of administrative divisions (i.e. nation, province, prefecture, county, town, and village). Each level has different types of urban or rural areas and ethnic autonomous regions, and there are differences between permanent residents and registered residents. Current population statistics are incomplete, inconsistent, non-standard and inaccurate. For data accuracy, completeness, and consistency, we use the data of the national and provincial statistical yearbooks to adjust the population of the whole country, provinces and prefectures in the preceding one year, and include the data of national health, resources and environments. Based on these big data, we assess the present situation and characteristics of population spatial distributions in China. Then we integrate elasticity factors such as economy, society, resource, environment, medicine, and health care, and reveal the main factors influencing population distribution changes through multiple semi-parametric and non-parametric correlation and regression methods. Taking the changes of main factors as inputs, we use several supervised time series methods to model and validate, and select the optimal model to predict the trend of population spatial distributions. Throughout the history, current situation and trend of population distributions, under the background and constraints of national strategies, this project proposal focuses on the elasticity factors, and puts forward policy recommendations on the regional pattern of population and urban and rural areas in China during the 14th Five-Year Plan period.

4/7205000283

疫情新常态下主体适应行为建模与经济发展预测

Modelling individual adaption and prediction of economic development under COVID-19 epidemic new normal

G0307.人口资源环境经济与劳动经济

G0302.行为经济与实验经济

健康与经济 主体适应行为 经济发展预测

health sciences and economics, individual adaption, prediction of economic development

智能化与万物互联时代导致数据安全、隐私保护以及数据的跨境流动具有了较大的不确定性，已经深刻地影响到经济主体的适应行为；全球范围内爆发的新型冠状病毒疫情发展的不确定性加剧了全球产业链和供应链的不确定性，也深远改变了经济主体的健康行为，并开始持续性对全球经济产生更大的不确定性影响。这种变化使得以研究理性经济主体为对象的管理和经济科学的研究面临更为复杂的个体行为和高度不确定性的复杂环境，给经济与管理学科的基础研究带来了巨大挑战，需要突破传统理论方式和范式的约束，瞄准具有变革性、可持续性的健康经济学科学前沿方向，探讨疫情新常态下的经济主体适应行为和目标约束下的内生决策过程及其对经济产出的影响。基于此本研究提出健康幸福导向的反馈-前馈自适应模型研究新冠疫情下经济主体如何实现健康幸福目标，并把这一目标纳入三组分经济发展预测模型，揭示在健康目标约束下，经济主体的自适应行为会如何影响经济产出，从而提出能够较好模拟未来经济发展的预测模型。

健康幸福导向的反馈-前馈自适应模型，以健康预期寿命和幸福感指数为目标，包括健康系数、健康投资占比等要素。健康和幸福融入异质经济主体投资和消费的各个方面，预期目标影响投资和消费即为前馈，当前健康和幸福的结果产出影响主体投入即为反馈，在前馈和反馈共同作用下主体主动适应当前和未来经济社会形势即为自适应。健康预期寿命，即无病生存时间；中国人平均8年以上带病生存，“健康中国行动”力促提高健康预期寿命。压力是中国伤残的主要危险因素，减少释放压力、提高幸福做为各项政策和经济主体的着力点。健康系数定义为医疗、运动、康养支出占总支出的比例，衡量消费结构和富裕程度；健康投资占比为社会保险、健康商业险、健康领域股票基金债券与直接投资占总投资的比例。通过神经网络和深度学习等人工智能算法评估自适应模型的灵敏度与稳健性、特异性与适用性、兼容度与扩展性。反馈-前馈自适应模型输出的健康预期寿命和幸福感指数融入三组分经济发展预测模型的要素之中。

三组分经济发展预测模型，包括微观因素、中观宏观因素和扰动因素；微观因素指的是经济发展的劳动、资本、土地、知识、技术、管理、数据等生产要素，中观宏观因素指的是产业结构、15岁以上人口平均受教育年限、总和生育率的组合指数，扰动因素指的是微观因素和中观宏观因素没有解释的与时间有关的差分整合自回归移动平均模型和健康幸福目标对模型的当期和长期扰动。微观生产要素回应新冠疫情防控新常态下产业链和供应链、金融、数据、地缘的不确定性，考量以国内大循环为主体、国内国际双循环相互促进的发展新格局对经济的影响。中观宏观因素，产业结构、15岁以上人口平均受教育年限、总和生育率均指数化（0-100）然后综合成经济柔韧性指数，回应经济学基础研究的复杂性，考量产业和人口结构对经济的影响；鉴于绝对量和相对量都对经济产生影响，申请人提出复合指数，以不同产业、地区绝对量中位数为基准用绝对量校正相对量或人均值（例如河南人均GDP校正值=名义值×河南人口÷省级人口中位数）。扰动因素，回应前期历史和偶然因素对即期和未来的影响。本研究具体刻画各组分的权重、各个组分中不同因素的权重、校正因子权重、有无健康幸福目标对该模型的扰动程度。

健康幸福导向的反馈-前馈自适应模型和三组分经济发展预测模型能够刻画异质微观主体的内生决策过程，研究其适应性决策行为和动态关联造就的宏观经济运行的特征和演化规律。本研究所提学术思想，第一是思想上原创，强调微观主体以健康幸福而不是资产、收支平衡为导向进行内生决策，宏观经济历经多轮弱平衡进行演化从而实现可持续发展目标；第二是方法上有创新，将复杂性科学、人工智能、行为科学与认知神经科学中的前馈-反馈环路引入微观主体适应性决策过程，提出三组分模型揭示基于微观主体适应行为的宏观经济演化规律；第三是研究基础和体系上科学，统计学习和健康经济学正在交叉融合，本项目基于微观主体抽样和调查数据进行决策行为和宏观经济的建模、验证和预测，采用经济柔韧性指标和三组分复合模型分别回应新冠疫情防控新常态下经济学基础理论面临的不确定性和复杂性。

本项目将经济科学同复杂性科学、数据科学、人工智能、行为科学与认知神经科学等进行跨学科跨领域的深度交叉融合，面向国家经济治理体系协调优化的重大需求，从复杂性科学的视角，突破经济科学的既有认知，探索和构建大数据和人工智能新技术驱动的复杂经济系统治理尤其是健康经济学新理论、新方法和新范式，推动中国经济科学前沿理论的变革性发展，为双循环发展提供理论基础，服务人类健康共同体和经济高质量可持续发展。

window-based Gradient Boosting Regression Tree

simpler machine learning baselines should not be dismissed and instead configured with more care to ensure the authenticity of the progression in field of time series forecasting.

https://arxiv.org/pdf/2101.02118.pdf

5/3215000278

基于类脑实时图计算与图数据库的生物数据安全共享

Security and sharing of biological data based on neuromorphic real-time graph computing and database

C2105  C0608

亚类说明： 指南引导类原创探索计划项目 附注说明： 未来生物技术

研究方向：生物大数据技术

数据为生物学要素之一，生物数据计算分析涉及数据获取、应用场景及主体权益等。单项数据技术和端系统建设进展显著，但是生态系统未建立；需要研发兼顾海量数据存储、深度计算和高并发通信的技术。群体学习结合边缘计算、基于区块链的对等网络和协调，实现数据安全共享。我们提炼人群数据各环节和联合研究的共性原理和知识，构建一套全流程处理机制；设计相邻哈希，研发类脑实时深度图数据库与图算法，能够高密度并发计算、动态剪枝，具备线性可扩展、高可视化；结合共归一化和群体学习等技术在端和云开展多样性队列研究和临床诊疗；通过混合模型构建人群数据的解释模型，完成各类疾病的风险预测和数据挖掘；端网云边群结合对计算存储、传输流动和网络资源进行统一智能管理、弹性调度，为用户提供多层面安全防护；形成千万以上人群数据研究和临床诊疗的标准和立法建议等，构建队列联合研究和临床诊疗联合体生态系统。从而推动人群健康研究范式转变及精准施策。

Data is an essential biological elements. Biological data processing needs to consider data acquisition, application scenarios and individual rights and interests. Significant progress has been made in the construction of single data technology and end systems, but the ecosystem has not yet been established. It is vital to study and develop the original technologies that simultaneously combine mass data storaging, deep computing, and high-parallel communicating. Swarm Learning combines edge computing, and blockchain based peer-to-peer network and coordination to realize data sharing under privacy and security conditions. We refine common principles and knowledge of population data and jointly research and build a whole-pipeline processing mechanism; design adjacency Hash and thus invent neuromorphic real-time deep graph database and graph algorithm, providing linear scalability, and high visualization via high-density concurrent computing power and algorithm, and dynamic trimming; use co-normalization, graph computing, swarm learning, and other technologies to carry out diverse cohort research and clinical diagnosis and treatment on node/end and cloud platforms; construct the explanation model of population data via mixed effect model, to perform the risk prediction and data mining of various diseases; combine node/terminal, network, cloud, edge, and swarm, to perform unified management, intelligent control and flexible scheduling of computing, storaging, transmission, flow and network resources, providing users with multi-level security protection; and form the standards and legislative suggestions for data research and clinical diagnosis and treatment of more than 10 million people, and build an ecosystem of cohort joint research and clinical diagnosis and treatment. Thus this project promotes the shift of research paradigm and accurate implementation of policies in population health.

中文关键词：

生物数据分析

生物大数据共享

图计算与图数据库

群体学习

共归一化

*英文关键词：

Biological Data Analysis

Biological Big Data Sharing

Graph Computing and Database

Swarm Learning

Co-normalization

主要研究领域：

生物数据学/生物数据信息智能技术（BDIIT）融合

• Biodatatics coined in June, 2020, is the fusion of and based on

statistical physics (scaling, renormalization) 2000

bioinformatics (variable) 2005

econometrics (structural equation model, time series) 2006

jurisprudence (right and duty, power and responsibility) 2006

biostatistics (inference, null hypothesis) 2008

analytical chemistry (NMR, XC-MS) 2008

epidemiology (population, time points) 2016

data science (small, big, fast, deep data) 2019

artificial intelligence (machine learning, pattern recognition, graph database) 2019

• Biodatatics is the discipline

that studies spatiotemporal diversity data (integration of variable, sample and time course) of organisms including microorganisms, plants, animals and human being, with methods of inference and significance test, and contents and goals of security, development, health and nutrition.

spatiotemporal biodiversity, -omics+epidemiology, prediction factors, full data, assembly biology, bioeconomics, fundamental theory

远缘融合统计物理学（标度和重正化群理论）、生物信息学（侧重变量）、生物统计学（样本推断总体、无效假设和混合模型）、流行病学（侧重样本和时点）、数据学（小、大和快、深和图数据，零试样本）、人工智能（模式识别、机器学习、知识表示与处理）与法理学，总结自己在发育健康营养多组学大数据方面的研究结果，于2020年6月提出生物数据学这一交叉学科。

生物数据学Biodatatics/BioDataScience是以生物（微生物、植物、动物、人）时空多样性［知识图谱atlas/knowledge graph］数据（变量、样本、时点）为对象，以无效假设、样本推断总体算法（人工智能、区块链、端网云边群）和显著性检验为手段，以安全发育健康营养为内容和目标的一门一级交叉学科。包括生物多样性、流行病学+组学、生物全数据学、生理与病理预测因素学、生物+经济社会、装配生物学和生物数据学基础理论等二级学科。

• Problem and solution

Population data: mass, scattered, multi-modal, cross-scale, mainly isolated, partially unsafe sharing; imbalanced demand and supply

Consider: data acquisition, application scenarios and individual rights and interests

Solution: simultaneously combine mass data storaging, deep computing, and high-parallel communicating

Diverse datasets, neuromorphic: adjacency Hash, similarity storage →knowledge-curated, scenario-targeted →native, concurrent, chain-style →high-visualization; DCSCE

• Data activity mechanism DCSCE
  

Diverse: integrated using native and neuromorphic methods (demi-schema & chained mode)

Coherent/Cohesive: Design, Acquisition, Analysis, Governance; APIs  & SDKs, re- & co-normalization, dynamic pruning

Secure and Shared/Common: anonymous and de-identified datasets with common feature are shared for joint contribution

Corresponding: parties of data own rights, undertake duties and attain interests with proportional principles

Ecological and Enabled: parties of data are better in health outcomes

全流程数据活动DCSCE机制

尊重多样（物理分散，多源多模态异构，不同质量）

环节衔接（不同数据环节之间和不同数据集之间留有接口，归一化标度、滤噪去重）(application programe interfaces)  (software development kits)

权责对应（数据是一种生产要数，按照比例原则承担相应的数据质量责任和获得收益）

生态使能（各数据参与者都能从中获益，促进健康结局）。

Please cite this blog if you refer to the above contents. 欢迎参考上述内容，共同推广生物数据学和数据活动DCSCE机制，构建数据生态系统。