Multi-modality machine learning predicting Parkinson's disease.
Multi-modality machine learning predicting Parkinson's disease.
复制标题
多模式机器学习预测帕金森氏病。
DOI：
10.1038/s41531-022-00288-w
复制
发表时间：
2022-04-01
期刊：
影响因子：
--
通讯作者：
Nalls MA
中科院分区：
文献类型：
--
作者：
Makarious MB;Leonard HL;Vitale D;Iwaki H;Sargent L;Dadu A;Violich I;Hutchins E;Saffo D;Bandres-Ciga S;Kim JJ;Song Y;Maleknia M;Bookman M;Nojopranoto W;Campbell RH;Hashemi SH;Botia JA;Carter JF;Craig DW;Van Keuren-Jensen K;Morris HR;Hardy JA;Blauwendraat C;Singleton AB;Faghri F;Nalls MA
个性化医疗承诺个性化疾病预测和治疗。机器学习（ML）和可用多模态数据的融合是向前发展的关键。我们在之前的工作基础上，提供了帕金森病（PD）风险的多模式预测，并系统地开发了一个使用基因组（一个自动化ML包）的模型，以改进PD的多组学预测，并在外部队列中得到验证。我们研究了顶级特征，构建了无假设的疾病相关网络，并研究了药物-基因相互作用。我们对来自帕金森病进展标志物（PPMI）的多模态数据进行了自动ML。在选择最佳算法后，使用所有PPMI数据对所选模型进行调优。该模型在帕金森病生物标志物计划（PDBP）数据集中得到验证。我们的初始模型显示，诊断PD的曲线下面积（AUC）为89.72%。然后对调整后的模型进行外部数据验证（PDBP， AUC为85.03%）。优化分类阈值提高了诊断预测的准确性和其他指标。最后，建立网络来识别PD特异性基因群落。结合数据模式优于单一生物标志物范式。UPSIT和PRS对模型的预测能力贡献最大，但它们的准确性被许多较小的效应转录本和风险snp所补充。我们的模型最适合于在健康登记处或生物库中确定要监测的大型个体群体，以便优先进行进一步的测试。这种方法允许复杂的预测模型可以被社区再现和访问，并且包、代码和结果都是公开可用的。
Personalized medicine promises individualized disease prediction and treatment. The convergence of machine learning (ML) and available multimodal data is key moving forward. We build upon previous work to deliver multimodal predictions of Parkinson’s disease (PD) risk and systematically develop a model using GenoML, an automated ML package, to make improved multi-omic predictions of PD, validated in an external cohort. We investigated top features, constructed hypothesis-free disease-relevant networks, and investigated drug–gene interactions. We performed automated ML on multimodal data from the Parkinson’s progression marker initiative (PPMI). After selecting the best performing algorithm, all PPMI data was used to tune the selected model. The model was validated in the Parkinson’s Disease Biomarker Program (PDBP) dataset. Our initial model showed an area under the curve (AUC) of 89.72% for the diagnosis of PD. The tuned model was then tested for validation on external data (PDBP, AUC 85.03%). Optimizing thresholds for classification increased the diagnosis prediction accuracy and other metrics. Finally, networks were built to identify gene communities specific to PD. Combining data modalities outperforms the single biomarker paradigm. UPSIT and PRS contributed most to the predictive power of the model, but the accuracy of these are supplemented by many smaller effect transcripts and risk SNPs. Our model is best suited to identifying large groups of individuals to monitor within a health registry or biobank to prioritize for further testing. This approach allows complex predictive models to be reproducible and accessible to the community, with the package, code, and results publicly available.
登录
查看更多内容
影响因子：
64.8
作者：
Green ED;Gunter C;Biesecker LG;Di Francesco V;Easter CL;Feingold EA;Felsenfeld AL;Kaufman DJ;Ostrander EA;Pavan WJ;Phillippy AM;Wise AL;Dayal JG;Kish BJ;Mandich A;Wellington CR;Wetterstrand KA;Bates SA;Leja D;Vasquez S;Gahl WA;Graham BJ;Kastner DL;Liu P;Rodriguez LL;Solomon BD;Bonham VL;Brody LC;Hutter CM;Manolio TA
通讯作者：
Manolio TA
影响因子：
4.6
作者：
Haynes, Winston A.;Tomczak, Aurelie;Khatri, Purvesh
通讯作者：
Khatri, Purvesh
影响因子：
3.5
作者：
Hill-Burns EM;Ross OA;Wissemann WT;Soto-Ortolaza AI;Zareparsi S;Siuda J;Lynch T;Wszolek ZK;Silburn PA;Mellick GD;Ritz B;Scherzer CR;Zabetian CP;Factor SA;Breheny PJ;Payami H
通讯作者：
Payami H
影响因子：
8.8
作者：
Fernandes, Hugo J. R.;Patikas, Nikolaos;Metzakopian, Emmanouil
通讯作者：
Metzakopian, Emmanouil
影响因子：
7.5
作者：
Geurts, P;Ernst, D;Wehenkel, L
通讯作者：
Wehenkel, L