Modeling Hippocampal Neurogenesis Using Human Pluripotent Stem CellsThe availability of human pluripotent stem cells (hPSCs) offers the opportunity to generate lineage-specific cells to investigate mechanisms of human diseases specific to brain regions. Here, we report a differentiation paradigm for hPSCs that enriches for hippocampal dentate gyrus (DG) granule neurons. This differentiation paradigm recapitulates the expression patterns of key developmental genes during hippocampal neurogenesis, exhibits characteristics of neuronal network maturation, and produces PROX1+ neurons that functionally integrate into the DG. Because hippocampal neurogenesis has been implicated in schizophrenia (SCZD), we applied our protocol to SCZD patient-derived human induced pluripotent stem cells (hiPSCs). We found deficits in the generation of DG granule neurons from SCZD hiPSC-derived hippocampal NPCs with lowered levels of NEUROD1, PROX1, and TBR1, reduced neuronal activity, and reduced levels of spontaneous neurotransmitter release. Our approach offers important insights into the neurodevelopmental aspects of SCZD and may be a promising tool for drug screening and personalized medicine.
Modeling Hippocampal Neurogenesis Using Human Pluripotent Stem Cells(Stem Cell Reports 2, 295–310, March 11, 2014) The Authors would like to acknowledge the JPB Foundation for the support of the work in this manuscript. Modeling Hippocampal Neurogenesis Using Human Pluripotent Stem CellsYu et al.Stem Cell ReportsFebruary 27, 2014In BriefGage and colleagues report a differentiation paradigm for human stem cells (ESCs and IPSCs) to generate hippocampal dentate gyrus (DG) granule neurons. This paradigm recapitulates characteristic gene expressions and produces functional neurons that integrate into the DG. Differentiation of schizophrenia patient-derived IPSCs revealed deficits in hippocampal neurogenesis from the NPC population and produced DG granule neurons with reduced spontaneous transmitter release. Full-Text PDF Open Access
Feature engineering in user's music preference predictionJianjun Xie, Scott Leishman, Tian Liang et al.|Knowledge Discovery and Data Mining|2011 The second track of this year's KDD Cup asked contestants to separate a user's highly rated songs from unrated songs for a large set of Yahoo! Music listeners. We cast this task as a binary classification problem and addressed it utilizing gradient boosted decision trees. We created a set of highly predictive features, each with a clear explanation. These features were grouped into five categories: hierarchical linkage features, track-based statistical features, user-based statistical features, features derived from the k-nearest neighbors of the users, and features derived from the k-nearest neighbors of the items. No music domain knowledge was needed to create these features. We demonstrate that each group of features improved the prediction accuracy of the classification model. We also discuss the top predictive features of each category in this paper.