Complex clinical, neurobiological, and molecular signatures of the longitudinal course of psychosis: leveraging comprehensive phenotyping, novel machine learning, and (epi)genomic approaches

PsyCourse的PC1项目的总体目标是鉴定主要精神疾病纵向病程的临床、神经生物学和分子遗传特征。由于它是基于临床研究小组241 （KFO 241）的成功工作，主要重点将放在精神分裂症（SZ）和双相情感障碍（BD）上。为了更好地解释精神病学表型之间的病因重叠，我们现在还将包括重度抑郁症（MDD）。在3年的时间里，我们将建立一个由1700名患者（700名SZ， 700名BD， 300名MDD）和500名对照者组成的深度表型队列，对这些患者进行至少4个时间点（0、6、12、18个月）的广泛表型和生物库（血液、DNA、RNA、血浆、血清）。通过KFO 241在德国范围内的招聘基础设施，这一独特的资源已经成为可能，因此将构成整个PsyCourse以人为本研究的核心资源。我们将应用先进的机器学习方法对这些队列中的基线和纵向表型信息进行分析，以确定新的目标表型过程和结果，以进行进一步的分子和统计分析。我们将评估个体化诊断分类和预后预测的可行性。大量的数据将为这些新的分析提供信息，如社会人口统计学、精神病理学、神经认知和基因组数据。我们项目的机器学习部分也将帮助确定患者在疾病过程的极端末端（最坏的和非常有利的过程）。我们将通过全基因组测序来询问这两组的完整基因组背景，以确定决定过程和结果的罕见变异的潜在参与。最后，利用尖端的microRNAnome测序技术，我们建议建立调控元件作为SZ病程的新生物标志物。因此，总的来说，我们的项目是PsyCourse的核心项目之一，因为它为联盟的临床研究奠定了基础，因为它描绘了新颖的、数据驱动的课程表型，这些表型被输入到PC1和PsyCourse中的其他分子分析中。

The overall goal of the PC1 project of PsyCourse is the identification of clinical, neurobiological, and molecular genetic signatures of the longitudinal course of major psychiatric disorders. As it is based upon the successful work of the Clinical Research Group 241 (KFO 241), a major focus will lie on schizophrenia (SZ) and bipolar disorder (BD). To better account for the etiological overlap between psychiatric phenotypes, we will now also include major depressive disorder (MDD). Over the course of 3 years, we will establish a deep-phenotyped cohort of 1,700 patients (700 SZ, 700 BD, 300 MDD) and 500 controls, for whom extensive phenotyping and biobanking (blood, DNA, RNA, plasma, serum) at at least 4 time points (0,6,12,18 months) will be available. This unique resource that has been made possible through KFO 241`s German-wide recruitment infrastructure will thus constitute the core resource for human-centered research across PsyCourse. We will apply advanced machine learning methods to the baseline and longitudinal phenotypic information in these cohorts to define novel target phenotypes of course and outcome for further molecular and statistical analyses. We will assess the feasibility of an individualized diagnostic classification and outcome prediction. A vast array of data will inform these novel analyses, such as socio-demographic, psychopathological, neurocognitive and genomic data. The machine-learning part of our project will also help define patients at the extreme ends of the course of disorder (worst vs. very favorable course). We will interrogate the complete genomic background of these two groups by whole-genomic sequencing to establish a potential involvement of rare variants determining course and outcome. Finally, using cutting-edge microRNAnome sequencing, we propose to establish regulatory elements as novel biomarkers for course in SZ. Thus, in its entirety, our project is one of the core projects within PsyCourse as it lays the foundation for the clinical research of the consortium and as it delineates novel, data-driven course phenotypes feeding into molecular analyses within PC1 and others within PsyCourse.