Modeling islet enhancers using deep learning identifies candidate causal variants at loci associated with T2D and glycemic traits.
Modeling islet enhancers using deep learning identifies candidate causal variants at loci associated with T2D and glycemic traits.
复制标题
DOI：
10.1073/pnas.2206612120
复制
发表时间：
2023-08-29
影响因子：
11.1
通讯作者：
Collins, Francis S.
中科院分区：
文献类型：
--
作者：
Hudaiberdiev, Sanjarbek;Taylor, D. Leland;Song, Wei;Narisu, Narisu;Bhuiyan, Redwan M.;Taylor, Henry J.;Tang, Xuming;Yan, Tingfen;Swift, Amy J.;Bonnycastle, Lori L.;Chen, Shuibing;Erdos, Michael R.;Ovcharenko, Ivan;Collins, Francis S.
识别与疾病相关的遗传变异的基因组和分子效应是将遗传关联研究的信号转化为对疾病原因的洞察的核心挑战。这种影响可以通过有针对性的功能研究来定义，但这些研究很难跨越数以千计的候选因果变异，这些候选因果变异通常由遗传关联研究确定。为了帮助解决这个问题，我们开发了一种方法来预测基因变异对增强子的影响。我们将这种方法应用于胰岛增强剂的建模，证明了该模型的准确性，并表明遗传变异对增强剂的预测效果可以帮助识别目标功能研究的候选因果变异。遗传关联研究已经确定了数百个与2型糖尿病(T2D)和相关特征相关的独立信号。尽管取得了这些成功，但识别遗传关联信号背后的特定因果变异仍然具有挑战性。在这项研究中，我们描述了一种深度学习(DL)方法来分析序列变体对增强子的影响。聚焦于胰岛，一种与T2D相关的组织，我们表明我们的模型学习了胰岛特异的转录因子(TF)调节模式，并可用于优先考虑候选因果变量。在101个与T2D和相关血糖性状相关的遗传信号中，多个变异出现在连锁不平衡中，我们的方法为每个关联信号指定一个因果变异，包括先前显示的改变胰岛相关细胞类型中报告活性的三个变异。对于另一个与血糖水平相关的信号，我们使用胰岛β细胞系对统计精细图谱中的所有候选因果变量进行生化测试，并显示等位基因对模型优先变量的Tf结合的影响的生化证据。为了帮助未来的研究，我们公开发布了我们的模型和胰岛增强子扰动分数，涉及约6700万个基因变体。我们预计，这项研究中提出的DL方法将加强对功能研究的候选因果变量的优先排序。
Identifying the genomic and molecular effects of disease-associated genetic variants is a central challenge in translating signals from genetic association studies to insights into the causes of disease. Such effects can be defined by targeted functional studies, but these studies are difficult to scale across the thousands of candidate causal variants routinely identified by genetic association studies. To help solve this problem, we developed a method to predict the effects of genetic variation on enhancers. We apply this method to model pancreatic islet enhancers, demonstrate that the model is accurate, and show that the predicted effects of genetic variants on enhancers can help identify candidate causal variants for targeted functional studies. Genetic association studies have identified hundreds of independent signals associated with type 2 diabetes (T2D) and related traits. Despite these successes, the identification of specific causal variants underlying a genetic association signal remains challenging. In this study, we describe a deep learning (DL) method to analyze the impact of sequence variants on enhancers. Focusing on pancreatic islets, a T2D relevant tissue, we show that our model learns islet-specific transcription factor (TF) regulatory patterns and can be used to prioritize candidate causal variants. At 101 genetic signals associated with T2D and related glycemic traits where multiple variants occur in linkage disequilibrium, our method nominates a single causal variant for each association signal, including three variants previously shown to alter reporter activity in islet-relevant cell types. For another signal associated with blood glucose levels, we biochemically test all candidate causal variants from statistical fine-mapping using a pancreatic islet beta cell line and show biochemical evidence of allelic effects on TF binding for the model-prioritized variant. To aid in future research, we publicly distribute our model and islet enhancer perturbation scores across ~67 million genetic variants. We anticipate that DL methods like the one presented in this study will enhance the prioritization of candidate causal variants for functional studies.
登录
查看更多内容
影响因子：
30.8
作者：
Lee D;Gorkin DU;Baker M;Strober BJ;Asoni AL;McCallion AS;Beer MA
通讯作者：
Beer MA
影响因子：
7
作者：
Kwasnieski JC;Fiore C;Chaudhari HG;Cohen BA
通讯作者：
Cohen BA
影响因子：
30.8
作者：
Mahajan A;Taliun D;Thurner M;Robertson NR;Torres JM;Rayner NW;Payne AJ;Steinthorsdottir V;Scott RA;Grarup N;Cook JP;Schmidt EM;Wuttke M;Sarnowski C;Mägi R;Nano J;Gieger C;Trompet S;Lecoeur C;Preuss MH;Prins BP;Guo X;Bielak LF;Below JE;Bowden DW;Chambers JC;Kim YJ;Ng MCY;Petty LE;Sim X;Zhang W;Bennett AJ;Bork-Jensen J;Brummett CM;Canouil M;Ec Kardt KU;Fischer K;Kardia SLR;Kronenberg F;Läll K;Liu CT;Locke AE;Luan J;Ntalla I;Nylander V;Schönherr S;Schurmann C;Yengo L;Bottinger EP;Brandslund I;Christensen C;Dedoussis G;Florez JC;Ford I;Franco OH;Frayling TM;Giedraitis V;Hackinger S;Hattersley AT;Herder C;Ikram MA;Ingelsson M;Jørgensen ME;Jørgensen T;Kriebel J;Kuusisto J;Ligthart S;Lindgren CM;Linneberg A;Lyssenko V;Mamakou V;Meitinger T;Mohlke KL;Morris AD;Nadkarni G;Pankow JS;Peters A;Sattar N;Stančáková A;Strauch K;Taylor KD;Thorand B;Thorleifsson G;Thorsteinsdottir U;Tuomilehto J;Witte DR;Dupuis J;Peyser PA;Zeggini E;Loos RJF;Froguel P;Ingelsson E;Lind L;Groop L;Laakso M;Collins FS;Jukema JW;Palmer CNA;Grallert H;Metspalu A;Dehghan A;Köttgen A;Abecasis GR;Meigs JB;Rotter JI;Marchini J;Pedersen O;Hansen T;Langenberg C;Wareham NJ;Stefansson K;Gloyn AL;Morris AP;Boehnke M;McCarthy MI
通讯作者：
McCarthy MI
影响因子：
16.6
作者：
Greenwald, William W.;Chiou, Joshua;Gaulton, Kyle J.
通讯作者：
Gaulton, Kyle J.
影响因子：
64.8
作者：
Karczewski, Konrad J;Francioli, Laurent C;MacArthur, Daniel G
通讯作者：
MacArthur, Daniel G