Dual-extraction modeling:A multi-modal deeplearning architecture for phenotypic prediction and functional gene mining of complex traits

Despite considerable advances in extracting crucial insights from bio-omics data to unravel the intricate mechanisms underlying complex traits,the absence of a universal multi-modal computational tool with robust interpretability for accurate phenotype prediction and identification of trait-associated genes remains a challenge.This study introduces the dual-extraction modeling(DEM)approach,a multi-modal deep-learning architecture designed to extract representative features from heterogeneous omics datasets,enabling the prediction of complex trait phenotypes.Through comprehensive benchmarking experiments,we demonstrate the efficacy of DEM in classification and regression prediction of complex traits.DEM consistently exhibits superior accuracy,robustness,generalizability,and flexibility.Notably,we establish its effectiveness in predicting pleiotropic genes that influence both flowering time and rosette leaf number,underscoring its commendable interpretability.In addition,we have developed user-friendly software to facilitate seamless utilization of DEM’s functions.In summary,this study presents a state-of-the-art approach with the ability to effectively predict qualitative and quantitative traits and identify functional genes,confirming its potential as a valuable tool for exploring the genetic basis of complex traits.

Efficient protocols and methods for high-throughput utilization of the Collaborative Cross mouse model for dissecting the genetic basis of complex traits

The Collaborative Cross(CC)mouse model is a next‐generation mouse genetic reference population(GRP)designated for a high‐resolution quantitative trait loci(QTL)mapping of complex traits during health and disease.The CC lines were generated from reciprocal crosses of eight divergent mouse founder strains composed of five classical and three wild‐derived strains.Complex traits are defined to be controlled by variations within multiple genes and the gene/environment interactions.In this article,we introduce and present variety of protocols and results of studying the host response to infectious and chronic diseases,including type 2 diabetes and metabolic diseases,body composition,immune response,colorectal cancer,susceptibility to Aspergillus fumigatus,Klebsiella pneumoniae,Pseudomonas aeruginosa,sepsis,and mixed infections of Porphyromonas gingivalis and Fusobacterium nucleatum,which were conducted at our laboratory using the CC mouse population.These traits are observed at multiple levels of the body systems,including metabolism,body weight,immune profile,susceptibility or resistance to the development and progress of infectious or chronic diseases.Herein,we present full protocols and step‐by‐step methods,implemented in our laboratory for the phenotypic and genotypic characterization of the different CC lines,mapping the gene underlying the host response to these infections and chronic diseases.The CC mouse model is a unique and powerful GRP for dissecting the host genetic architectures underlying complex traits,including chronic and infectious diseases.

基于田间表型性状的火龙果优异种质资源评价研究

为了科学确定火龙果优异种质资源评价标准,以农业农村部南宁火龙果种质资源圃内保存的种质资源为试材,从植物学特征、开花结果特性、果实品质性状、抗性等方面比较分析及综合评价。结果表明,筛选确定18项优良、特异种质资源评价性状,其中优良种质的性状包括生长势、柱头与花药相对位置、单果质量、果肉质地、果肉风味、可溶性固形物含量、开花结果批次、着果率、商品果率、产量、溃疡病抗性、抗寒性、抗热性共13项,特异种质的性状包括花瓣颜色、茎蔓上刺量、果皮颜色、果肉颜色、单果质量、可溶性固形物含量、开花结果批次、产量、溃疡病抗性共9项。筛选出丰产性高的种质69份,大果种质5份,高糖度种质22份,果皮颜色优异的种质5份,果肉颜色特异的种质14份,无草腥味且风味较好的白肉种质4份,高抗溃疡病种质1份。

Complex genetic architecture underlying the plasticity of maize agronomic traits

Phenotypic plasticity is the ability of a given genotype to produce multiple phenotypes in response to changing environmental conditions.Understanding the genetic basis of phenotypic plasticity and establishing a predictive model is highly relevant to future agriculture under a changing climate.Here we report findings on the genetic basis of phenotypic plasticity for 23 complex traits using a diverse maize population planted at five sites with distinct environmental conditions.We found that latituderelated environmental factors were the main drivers of across-site variation in flowering time traits but not in plant architecture or yield traits.For the 23 traits,we detected 109 quantitative trait loci(QTLs),29 for mean values,66 for plasticity,and 14 for both parameters,and 80%of the QTLs interacted with latitude.The effects of several QTLs changed in magnitude or sign,driving variation in phenotypic plasticity.We experimentally validated one plastic gene,ZmTPS14.1,whose effect was likely mediated by the compensation effect of ZmSPL6 from a downstream pathway.By integrating genetic diversity,environmental variation,and their interaction into a joint model,we could provide site-specific predictions with increased accuracy by as much as 9.9%,2.2%,and 2.6%for days to tassel,plant height,and ear weight,respectively.This study revealed a complex genetic architecture involving multiple alleles,pleiotropy,and genotype-byenvironment interaction that underlies variation in the mean and plasticity of maize complex traits.It provides novel insights into the dynamic genetic architecture of agronomic traits in response to changing environments,paving a practical way toward precision agriculture.

Future impact of integrated high-throughput methylome analyses on human health and disease

A spate of high-powered genome-wide association studies （GWAS） have recently identified numerous single-nucleotide polymorphisms （SNPs） robustly linked with complex disease. Despite interrogating the majority of common human variation, these SNPs only account for a small proportion of the phenotypic variance, which suggests genetic factors are acting in concert with non-genetic factors. Although environmental measures are logical covariants for genotype-phenotype investigations, another non-genetic intermediary exists： epigenetics. Epigenetics is the analysis of somatically-acquired and, in some cases, transgenerationally inherited epigenetic modifications that regulate gene expression, and offers to bridge the gap between genetics and environment to understand phenotype. The most widely studied epigenetic mark is DNA methylation. Aberrant methylation at gene promoters is strongly implicated in disease etiology, most notably cancer. This review will highlight the importance of DNA methylation as an epigenetic regulator, outline techniques to characterize the DNA methylome and present the idea of reverse phenotyping, where multiple layers of analysis are integrated at the individual level to create personalized digital phenotypes and, at a phenotype level, to identify novel molecular signatures of disease.

基于结构方程模型的疾病性状相关基因的识别

在复杂疾病中,疾病性状受许多基因控制,是多个基因共同作用的结果。但是多个基因以及与性状之间的相互作用关系都是庞大且复杂的,因此很难量化多个基因的共同作用从而找到疾病性状相关基因。论文提出了运用结构方程模型来对基因和疾病性状进行关联分析研究,依次来识别与疾病性状相关的基因。结构方程模型对于估计多个变量及复杂疾病之间的结构关系上具有优势,是衡量多个基因的综合作用,理清基因与疾病性状之间的关联关系的一种很好的方法。通过在真实数据上应用论文的方法,论文找到了一组和疾病性状相关的基因,并且通过相关分析验证了它们的综合作用关系。

从QTL到QTG的路还有多远?

植物大多数重要的经济性状都是数量性状,人们对许多植物进行了数量性状基因座(QTL)的研究,并取得了长足的发展。文章详尽地分析了数量性状表型与基因型的复杂关系,介绍了当前QTL研究领域里的几种精细作图策略。讨论了当前挖掘控制目标性状QTL基因的研究过程中存在的困难和问题,提出几个有待发展的研究方向,并展望了该领域的发展前景。因目前的QTL仍然是一个相当大的染色体区段,往往含有多个候选基因。文章就怎样从QTL粗放位点研究进一步发展到数量性状基因(quantitativetraitgene,QTG)水平上的变异,再从QTG到相应于基因内多态性的数量性状核苷酸(quantitativetraitnucleotides,QTN),提出了一些见解。来迎接后基因组时代数量遗传领域的挑战。

全基因组关联分析的进展与反思

全基因组关联分析(genomewide association study,GWAS)是应用人类基因组中数以百万计的单核苷酸多态性(single nucleotide polymorphism,SNP)为标记进行病例-对照关联分析,以期发现影响复杂性疾病发生的遗传特征的一种新策略。近年来,随着人类基因组计划和基因组单倍体图谱计划的实施,人们已通过GWAS方法发现并鉴定了大量与人类性状或复杂性疾病关联的遗传变异,为进一步了解控制人类复杂性疾病发生的遗传特征提供了重要的线索。然而,由于造成复杂性疾病/性状的因素较多,而且GWAS研究系统较为复杂,因此目前GWAS本身亦存在诸多的问题。本文将从研究方式、研究对象、遗传标记,以及统计分析等方面,探讨GWAS的研究现状以及存在的潜在问题,并展望GWAS今后的发展方向。

鸡MHC与传染性疾病遗传抗性的相关性研究进展

鸡是我国主要的家禽品种,抗病分子育种在鸡的疾病尤其是传染病控制中有着重要地位,抗性基因选择是其技术关键。鸡主要组织相容性复合体(MHC)基因具有高度多态性,与多种传染性疾病抗性紧密相关,受到家禽育种专家的高度关注。文章介绍国外有关鸡MHC与传染性疾病抗性的相关性及抗性基因研究进展,并展望其在鸡抗病育种中的应用前景。

全基因组关联研究中的交互作用研究现状

利用高密度单核苷酸多态(Single nucleotide polymorphism,SNP)标记在全基因组范围内检测影响复杂疾病/性状的染色体区段或基因,已经成为目前遗传学领域新的突破点之一。在全基因组关联研究(Genome-wide association study,GWAS)取得大量成果之后,研究者们对在全基因范围内研究交互作用产生了极大的热情。近几年,对交互作用的研究,无论是在方法的研发、实际的应用以及统计学上的交互向生物学上的交互转化,还是在信息组学的整合,都呈现快速发展的趋势。已有很多策略和方法被尝试用于进行全基因组交互作用分析,这些研究推动了对复杂疾病/性状遗传机制的进一步认识。基于目前全基因组交互分析所采用的各类数据处理方法的理论与算法的异同,文章拟对目前使用较为广泛的回归类方法、机器学习方法、贝叶斯模型法、SNP筛选类方法和基于并行程序的方法等5类方法加以评述,着重介绍了这些方法的算法原理、计算效率以及差别之处,以期能够为相关领域的研究者提供参考。

全基因组关联研究现状

在过去的5年中,全基因组关联研究(Genome-wide association study,GWAS)方法已被证明是研究复杂疾病和性状遗传易感变异的一种有效手段。目前,各国科学家在多种复杂疾病和性状中开展了大量的GWAS,对肿瘤、糖尿病、心脏病、神经精神疾病、自身免疫及免疫相关疾病等复杂疾病以及一些常见性状(如身高、体重、血脂、色素等)的遗传易感基因研究取得了重大成果。截止到2010年9月11日,运用GWAS开展了对近200种复杂疾病/性状的研究,发现了3000多个疾病相关的遗传变异。文章就GWAS的发展及其在复杂疾病/性状中的应用做一综述。

全基因组关联研究的深度分析策略

2005年至今,全基因组关联研究(Genome-wide association study,GWAS)发现了大量复杂疾病/性状相关变异。近来,科学家们关注的焦点又集中在了如何利用GWAS数据进行深入分析,期待发现更多复杂疾病/性状的易感基因。一些新的策略和方法已经被尝试应用到复杂疾病/性状GWAS的后续研究中,例如深入分析GWAS数据;鉴定新的复杂疾病/性状易感基因/位点;国际合作和Meta分析;易感区域精细定位及测序;多种疾病共同易感基因研究;以及基因型填补,基于通路的关联分析,基因-基因、基因-环境交互作用和上位研究等。这些策略和方法的应用弥补了经典GWAS的一些不足之处,进一步推动了人类对复杂疾病/性状遗传机制的认识。文章对上述研究的策略、方法以及所面临的问题和挑战进行了综述,为读者描绘了GWAS后期工作的一个简要框架。

复杂疾病基因鉴定的计算生物学方法

过去20多年复杂疾病易感基因鉴定的主要方法是连锁分析和关联研究。因为连锁分析确定的数量性状位点通常很宽,加之对区域内大部分基因的功能以及基因功能和疾病之间联系的认识十分有限,所以从数量性状位点到基因的识别是一个挑战。近年来发展了一些利用公共数据库的信息预测疾病易感基因的计算生物学方法。文章简要介绍了DGP、GeneSeeker、Prioritizer、PROSPECTR and SUSPECTS及Endeavor5种计算生物学方法的基本原理,以2型糖尿病/肥胖和骨质疏松症易感基因的预测为例说明它们的应用方法,并讨论了这些方法的局限及应用前景。

全基因组关联分析的研究现状及对数据科学的挑战

全基因组关联分析(GWAS),是通过考察全基因组范围DNA变异的单核苷酸多态性(SNP),挖掘影响复杂疾病等的表型性状(如疾病、癌症、身高等)的SNP的计算方法,以期为疾病/表型的分子生物发现、生物机理分析、分子靶向药物研究、疾病早期风险预测和个性化治疗等提供科学依据.目前的方法多以统计学、机器学习和深度学习、智能优化等等及其它们的组合为基础,并已取得可喜成绩,但仍有许多无法复现的关联的例子,正如Ioannidis 2005年在国际知名刊物PLoS Medicine上发表、至今已被引用6 600多次的论文中所说"大部分的研究发现是错的".文章认为,这是因为其核心问题仍未解决,尤其是到底要从数据中挖掘出什么和统计重要性在什么情况下具有科学重要性,以及科学重要性是否可以科学定义等,这些都是GWAS对数据科学的严峻挑战.

两种常用数量性状连锁分析方法的原理和进展

在复杂性状疾病的家系连锁研究中,Haseman Elston回归分析和方差组成模型是常用的两种数量性状连锁分析方法。前者主要针对同胞对的性状值差或和的平方进行回归分析;后者引用方差组成模型,将数量性状分解为遗传方差和环境方差,可估计二者对表型的影响。两种方法可应用于同胞对、核心家系或扩展家系,定位数量性状基因座。此文对这两种模型的原理、算法及其进展进行了综述,并给出了常用的统计软件包。

全基因组基因-基因相互作用研究现状

复杂疾病目前正在全球范围流行,极大地影响人类的健康。研究发现,复杂疾病的性状受到多个位点的相互作用影响。目前的全基因组关联分析(Genome-wide association study,GWAS)仅仅解析单个SNP位点对疾病易感性的贡献,单纯依靠这一种策略并不能在寻找复杂疾病的病因上得到根本性的突破。基因-基因相互作用可能是复杂疾病致病的主要因素之一。针对这一点,科学家已经提出了一些检验基因相互作用的算法,包括惩罚logistic回归模型、多因子降维(Multifactor dimensional reduction)、集合关联法(Set-association approach)、贝叶斯网络(Bayesian networks)、随机森林法等。文章首先对目前这些方法做了综述,并指出了其中的不足,包括计算复杂度太高、假设驱动、数据会过度拟合、对低维数据不敏感等,进而简述了一种由笔者所在实验室开发的基于GPU的研究基因相互作用的算法,该算法复杂度低,不需要任何假设,没有边际效应,有很好的稳定性,速度快,适用于进行全基因组范围内的基因-基因相互作用计算。

Using Composite Phenotypes to Reveal Hidden Physiological Heterogeneity in High-Altitude Acclimatization in a Chinese Han Longitudinal Cohort

Altitude acclimatization is a human physiological process of adjusting to the decreased oxygen availability.Since several physiological processes are involved and their correlations are complicated,the analyses of single traits are insufficient in revealing the complex mechanism of high-altitude acclimatization.In this study,we examined these physiological responses as the composite phenotypes that are represented by a linear combination of physiological traits.We developed a strategy that combines both spectral clustering and partial least squares path modeling(PLSPM)to define composite phenotypes based on a cohort study of 883 Chinese Han males.In addition,we captured 14 composite phenotypes from 28 physiological traits of high-altitude acclimatization.Using these composite phenotypes,we applied k-means clustering to reveal hidden population physiological heterogeneity in high-altitude acclimatization.Furthermore,we employed multivariate linear regression to systematically model(Models 1 and 2)oxygen saturation(SpO_(2))changes in high-altitude acclimatization and evaluated model fitness performance.Composite phenotypes based on Model 2 fit better than single trait-based Model 1 in all measurement indices.This new strategy of using composite phenotypes may be potentially employed as a general strategy for complex traits research such as genetic loci discovery and analyses of phenomics.

Tsc2 mutation induces renal tubular cell nonautonomous disease

TSC renal cystic disease is poorly understood and has no approved treatment.In a new principal cell-targeted murine model of Tsc cystic disease,the renal cystic epithelium is mostly composed of type A intercalated cells with an intact Tsc2 gene confirmed by sequencing,although these cells exhibit a Tsc-mutant disease phenotype.We used a newly derived targeted murine model in lineage tracing and extracellular vesicle(EV)characterization experiments and a cell culture model in EV characterization and cellular induction experiments to understand TSC cystogenesis.Using lineage tracing experiments,we found principal cells undergo clonal expansion but contribute very few cells to the cyst.We determined that cystic kidneys contain more interstitial EVs than noncystic kidneys,excrete fewer EVs in urine,and contain EVs in cyst fluid.Moreover,the loss of Tsc2 gene in EV-producing cells greatly changes the effect of EVs on renal tubular epithelium,such that the epithelium develops increased secretory and proliferative pathway activity.We demonstate that the mTORC1 pathway activity is independent form the EV production,and that the EV effects for a single cell line can vary significantly.TSC cystogenesis involves significant contribution from genetically intact cells conscripted to the mutant phenotype by mutant cell derived EVs.