<p>Gene regulatory networks (GRNs) provide a systems-level orchestration of an organism’s genome encoded anatomy. As biological networks are revealed, they continue to answer many questions including knowledge of how GRNs control morphogenetic movements and how GRNs evolve. Morphogenesis is a complex orchestration of movements by cells that are specified early in development. </p><p> The activation of an upstream GRN is crucial in order to orchestrate downstream morphogenetic events. In the sea urchin, activation of the endomesoderm GRN occurs after the asymmetric 4th cleavage. Embryonic asymmetric cell divisions often are accompanied by differential segregation of fate-determinants into one of two daughter cells. That asymmetric cleavage of the sea urchin micromeres leads to a differential animal-vegetal (A/V) nuclear accumulation of cell fate determinants, β-Catenin and SoxB1. Β-Catenin protein is localized into the nuclei of micromeres and activates the endomesoderm gene regulatory network, while SoxB1 is excluded from micromeres and enters the nucleus of the macromeres, the large progeny of the unequal 4th cleavage. Although nuclear localization of β-Catenin and SoxB1 shows dependence on the asymmetric cleavage, the mechanics behind the asymmetrical division has not been demonstrated. In Chapter 3, we show that micromere formation requires the small RhoGTPase, Cdc42 by directing the apical/basal orientation of the mitotic spindle at the apical cortex. By attenuating or augmenting sea urchin Cdc42 function, micromere divisions became defective and failed to correctly localize asymmetrically distributed determinants. As a consequence, cell fates were altered and multiple A/V axes were produced resulting in a “Siamese-twinning” phenotype that occurred with increasing frequency depending on the quantitative level of perturbation. Our findings show that Cdc42 plays a pivotal role in the asymmetric division of the micromeres, endomesoderm fate-determinant segregation, and A/V axis formation.</p><p> This dissertation also characterizes, at high resolution, the repertoire of cellular movements contributing to three different morphogenetic processes of sea urchin development: the elongation of gut, the formation of the primary mouth, and the migration of the small micromeres (the presumptive primordial germ cells) in the sea urchin, Lytechinus variegatus. Descriptive studies of the cellular processes during the different morphogenetic movements allow us to begin investigating their molecular control. </p><p>In Chapter 4, we dissected the series of complex events that coordinate gut and mouth morphogenesis. Until now, it was thought that lateral rearrangement of endoderm cells by convergent extension was the main contributor to sea urchin archenteron elongation and that cell divisions were minimal during elongation. We performed cell transplantations to live image and analyze a subset of labeled endoderm cells at high-resolution in the optically clear sea urchin embryo. We found that the endomesoderm cells that initially invaginate into the sea urchin blastocoel remained contiguous throughout extension, so that, if convergent extension were present, it was not a major contributor to elongation. We also found a prevalence of cell divisions throughout archenteron elongation that increased the number of cells within the gut linearly over time; however, we showed that the proliferation did not contribute to growth, and their spindle orientations were randomized during divisions and therefore did not selectively contribute to the final gut length. When cell divisions were inhibited, we saw no difference in the ability of the cells within the gut to migrate in order to elongate. Also in Chapter 4, we describe our observations of the cell biological processes underlying primary mouth formation at the end of gastrulation. Using time-lapse microscopy, photo-convertible Kaede, and an assay of the basement membrane remodeling, we describe a sequential orchestration of events that leads to the fusion of the oral ectoderm and the foregut endoderm. Our work characterizes, at higher resolution than previously recorded, the temporal sequence and repertoire of the cellular movements contributing to the length of the sea urchin larval gut and tissue fusion with the larval primary mouth.</p><p> In Chapter 5, the migration of the small micromeres to the coelomic pouches in the sea urchin embryo provides an exceptional model for understanding the genomic regulatory control of morphogenesis. An assay using the robust homing potential of these cells reveals a “coherent feed-forward” transcriptional subcircuit composed of Pax6, Six3, Eya, and Dach1 that is responsible for the directed homing mechanism of these multipotent progenitors. The linkages of that circuit are strikingly similar to a circuit involved in retinal specification in Drosophila suggesting that systems-level tasks can be highly conserved even though the tasks drive unrelated processes in different animals.</p><p> The sea urchin gene regulatory network (GRN) describes the cell fate specification of the developing embryo; however, the GRN does not describe specific cell biological events driving the three distinct sequences of cell movements. Our ability to connect the GRN to the morphogenetic events of gastrulation, primary mouth formation, and small micromere migration will provide a framework for characterizing these remarkable sequences of cell movements in the simplest of deuterostome models at an unprecedented scale.</p> / Dissertation

The objective of the field of Synthetic Biology is to implement novel functionalities in a biological context or redesign existing biological systems. To achieve this, it employs tried and tested engineering principles, such as standardisation and the design-build-test cycle. A crucial part of this process is the convergence of modelling and experiment. The aim of this thesis is to improve the design principles employed by Synthetic Biology in the context of Gene Regulatory Networks (GRNs). Small Ribonucleic Acids (sRNAs), in particular, are focussed on as a mechanism for post-transcriptional expression regulation, as they present great potential as a tool to be harnessed in GRNs. Modelling sRNA regulation and its interaction with its associated chaperone Host-Factor of Bacteriophage Q&beta; (Hfq) is investigated. Inclusion of Hfq is found to be necessary in stochastic models, but not in deterministic models. Secondly, feedback is core to the thesis, as it presents a means to scale-up designed systems. A linear design framework for GRNs is then presented, focussing on Transcription Factor (TF) interactions. Such frameworks are powerful as they facilitate the design of feedback. The framework supplies a block diagram methodology for visualisation and analysis of the designed circuit. In this context, phase lead and lag controllers, well-known in the context of Control Engineering, are presented as genetic motifs. A design example, employing the genetic phase lag controller, is then presented, demonstrating how the developed framework can be used to design a genetic circuit. The framework is then extended to include sRNA regulation. Four GRNs, demonstrating the simplest forms of genetic feedback, are then modelled and implemented. The feedback occurs at three different levels: autoregulation, through an sRNA and through another TF. The models of these GRNs are inspired by the implemented biological topologies, focussing on steady state behaviour and various setups. Both deterministic and stochastic models are studied. Dynamic responses of the circuits are also briefly compared. Data is presented, showing good qualitative agreement between models and experiment. Both culture level data and cell population data is presented. The latter of these is particularly useful as the moments of the distributions can be calculated and compared to results from stochastic simulation. The fit of a deterministic model to data is attempted, which results in a suggested extension of the model. The conclusion summarises the thesis, stating that modelling and experiment are in good qualitative agreement. The required next step is to be able to predict behaviour quantitatively.

Many diseases, such as cardiovascular diseases, cancer and diabetes, originate from several malfunctions in biological systems. The human body is regulated by a wide range of biological systems, composed of biological entities interacting in complex networks, responsible for carrying out specific functions. Some parts of the networks, such as hubs serving as master regulators, are more important for maintaining a function. To find the cause of diseases, where hubs are possible disease regulators, it is critical to know the structure of these biological systems. Such structures can be reverse engineered from high-throughput data with measured levels of biological entities. However, the complexity of biological systems makes inferring their structure a complicated task, demanding the use of computational methods, called network inference methods. Today, many network inference methods have been developed, that predicts the interactions of biological networks, with varying degree of success. In the DREAM5 challenge 35 network inference methods were evaluated on how well interactions in gene regulatory networks (GRNs) were predicted. Herein, in contrast to the DREAM5 challenge, we have evaluated network inference methods’ ability to predict hubs in GRNs. In accordance with the DREAM5 challenge, different methods performed the best on different data sets. Moreover, we discovered that network inference methods were not able to identify hubs from groups of similarly expressed genes. Also, we noticed that hubs in GRNs had a distinct expression in the data, leading to the development of a new method (the PCA method) for the prediction of hubs. Furthermore, the DREAM5 challenge showed that community predictions, combining the predictions from many network inference methods, resulted in more robust predictions of interactions. Herein, the community approach was applied on predicting hubs, with the conclusion that community predictions is the more robust approach. However, we also concluded that it was enough to combine 6-7 network inference methods to achieve robust predictions of hubs.

Nesta tese estamos especificamente interessados no problema de engenharia re- versa de redes regulatórias genéticas a partir de dados de pós-genômicos, um grande desafio na área de Bioinformática. Redes regulatórias genéticas são complexos cir- cuitos biológicos responsáveis pela regulação do nível de expressão dos genes, desem- penhando assim um papel fundamental no controle de inúmeros processos celulares, incluindo diferenciação celular, ciclo celular e metabolismo. Decifrar a estrutura destas redes é crucial para possibilitar uma maior compreensão à nível de sistema do desenvolvimento e comportamento dos organismos, e eventualmente esclarecer os mecanismos de doenças causados pela desregulação dos processos acima mencio- nados. Devido ao expressivo aumento da disponibilidade de dados experimentais de larga escala e da grande dimensão e complexidade dos sistemas biológicos, métodos computacionais têm sido ferramentas essenciais para viabilizar esta investigação. No entanto, seu desempenho ainda é bastante deteriorado por importantes desafios com- putacionais e biológicos impostos pelo cenário. Em particular, o ruído e esparsidade inerentes aos dados biológicos torna este problema de inferência de redes um difícil problema de otimização combinatória, para o qual métodos computacionais dispo- níveis falham em relação à exatidão e robustez das predições. Esta tese tem como objetivo investigar o uso de técnicas de ensemble learning como forma de superar as limitações existentes e otimizar o processo de inferência, explorando a diversidade entre um conjunto de modelos. Com este intuito, desenvolvemos métodos computa- cionais tanto para gerar redes diversificadas, como para combinar estas predições em uma solução única (solução ensemble ), e aplicamos esta abordagem a uma série de cenários com diferentes fontes de diversidade a fim de compreender o seu potencial neste contexto específico. Mostramos que as soluções propostas são competitivas com algoritmos tradicionais deste campo de pesquisa e que melhoram nossa capa- cidade de reconstruir com precisão as redes regulatórias genéticas. Os resultados obtidos para a inferência de redes de regulação transcricional e pós-transcricional, duas camadas adjacentes e complementares que compõem a rede de regulação glo- bal, tornam evidente a eficiência e robustez da nossa abordagem, encorajando a consolidação de ensemble learning como uma metodologia promissora para decifrar a estrutura de redes regulatórias genéticas. / In this thesis we are concerned about the reverse engineering of gene regulatory networks from post-genomic data, a major challenge in Bioinformatics research. Gene regulatory networks are intricate biological circuits responsible for govern- ing the expression levels (activity) of genes, thereby playing an important role in the control of many cellular processes, including cell differentiation, cell cycle and metabolism. Unveiling the structure of these networks is crucial to gain a systems- level understanding of organisms development and behavior, and eventually shed light on the mechanisms of diseases caused by the deregulation of these cellular pro- cesses. Due to the increasing availability of high-throughput experimental data and the large dimension and complexity of biological systems, computational methods have been essential tools in enabling this investigation. Nonetheless, their perfor- mance is much deteriorated by important computational and biological challenges posed by the scenario. In particular, the noisy and sparse features of biological data turn the network inference into a challenging combinatorial optimization prob- lem, to which current methods fail in respect to the accuracy and robustness of predictions. This thesis aims at investigating the use of ensemble learning tech- niques as means to overcome current limitations and enhance the inference process by exploiting the diversity among multiple inferred models. To this end, we develop computational methods both to generate diverse network predictions and to combine multiple predictions into an ensemble solution, and apply this approach to a number of scenarios with different sources of diversity in order to understand its potential in this specific context. We show that the proposed solutions are competitive with tra- ditional algorithms in the field and improve our capacity to accurately reconstruct gene regulatory networks. Results obtained for the inference of transcriptional and post-transcriptional regulatory networks, two adjacent and complementary layers of the overall gene regulatory network, evidence the efficiency and robustness of our approach, encouraging the consolidation of ensemble systems as a promising methodology to decipher the structure of gene regulatory networks.

Nesta tese estamos especificamente interessados no problema de engenharia re- versa de redes regulatórias genéticas a partir de dados de pós-genômicos, um grande desafio na área de Bioinformática. Redes regulatórias genéticas são complexos cir- cuitos biológicos responsáveis pela regulação do nível de expressão dos genes, desem- penhando assim um papel fundamental no controle de inúmeros processos celulares, incluindo diferenciação celular, ciclo celular e metabolismo. Decifrar a estrutura destas redes é crucial para possibilitar uma maior compreensão à nível de sistema do desenvolvimento e comportamento dos organismos, e eventualmente esclarecer os mecanismos de doenças causados pela desregulação dos processos acima mencio- nados. Devido ao expressivo aumento da disponibilidade de dados experimentais de larga escala e da grande dimensão e complexidade dos sistemas biológicos, métodos computacionais têm sido ferramentas essenciais para viabilizar esta investigação. No entanto, seu desempenho ainda é bastante deteriorado por importantes desafios com- putacionais e biológicos impostos pelo cenário. Em particular, o ruído e esparsidade inerentes aos dados biológicos torna este problema de inferência de redes um difícil problema de otimização combinatória, para o qual métodos computacionais dispo- níveis falham em relação à exatidão e robustez das predições. Esta tese tem como objetivo investigar o uso de técnicas de ensemble learning como forma de superar as limitações existentes e otimizar o processo de inferência, explorando a diversidade entre um conjunto de modelos. Com este intuito, desenvolvemos métodos computa- cionais tanto para gerar redes diversificadas, como para combinar estas predições em uma solução única (solução ensemble ), e aplicamos esta abordagem a uma série de cenários com diferentes fontes de diversidade a fim de compreender o seu potencial neste contexto específico. Mostramos que as soluções propostas são competitivas com algoritmos tradicionais deste campo de pesquisa e que melhoram nossa capa- cidade de reconstruir com precisão as redes regulatórias genéticas. Os resultados obtidos para a inferência de redes de regulação transcricional e pós-transcricional, duas camadas adjacentes e complementares que compõem a rede de regulação glo- bal, tornam evidente a eficiência e robustez da nossa abordagem, encorajando a consolidação de ensemble learning como uma metodologia promissora para decifrar a estrutura de redes regulatórias genéticas. / In this thesis we are concerned about the reverse engineering of gene regulatory networks from post-genomic data, a major challenge in Bioinformatics research. Gene regulatory networks are intricate biological circuits responsible for govern- ing the expression levels (activity) of genes, thereby playing an important role in the control of many cellular processes, including cell differentiation, cell cycle and metabolism. Unveiling the structure of these networks is crucial to gain a systems- level understanding of organisms development and behavior, and eventually shed light on the mechanisms of diseases caused by the deregulation of these cellular pro- cesses. Due to the increasing availability of high-throughput experimental data and the large dimension and complexity of biological systems, computational methods have been essential tools in enabling this investigation. Nonetheless, their perfor- mance is much deteriorated by important computational and biological challenges posed by the scenario. In particular, the noisy and sparse features of biological data turn the network inference into a challenging combinatorial optimization prob- lem, to which current methods fail in respect to the accuracy and robustness of predictions. This thesis aims at investigating the use of ensemble learning tech- niques as means to overcome current limitations and enhance the inference process by exploiting the diversity among multiple inferred models. To this end, we develop computational methods both to generate diverse network predictions and to combine multiple predictions into an ensemble solution, and apply this approach to a number of scenarios with different sources of diversity in order to understand its potential in this specific context. We show that the proposed solutions are competitive with tra- ditional algorithms in the field and improve our capacity to accurately reconstruct gene regulatory networks. Results obtained for the inference of transcriptional and post-transcriptional regulatory networks, two adjacent and complementary layers of the overall gene regulatory network, evidence the efficiency and robustness of our approach, encouraging the consolidation of ensemble systems as a promising methodology to decipher the structure of gene regulatory networks.

Nesta tese estamos especificamente interessados no problema de engenharia re- versa de redes regulatórias genéticas a partir de dados de pós-genômicos, um grande desafio na área de Bioinformática. Redes regulatórias genéticas são complexos cir- cuitos biológicos responsáveis pela regulação do nível de expressão dos genes, desem- penhando assim um papel fundamental no controle de inúmeros processos celulares, incluindo diferenciação celular, ciclo celular e metabolismo. Decifrar a estrutura destas redes é crucial para possibilitar uma maior compreensão à nível de sistema do desenvolvimento e comportamento dos organismos, e eventualmente esclarecer os mecanismos de doenças causados pela desregulação dos processos acima mencio- nados. Devido ao expressivo aumento da disponibilidade de dados experimentais de larga escala e da grande dimensão e complexidade dos sistemas biológicos, métodos computacionais têm sido ferramentas essenciais para viabilizar esta investigação. No entanto, seu desempenho ainda é bastante deteriorado por importantes desafios com- putacionais e biológicos impostos pelo cenário. Em particular, o ruído e esparsidade inerentes aos dados biológicos torna este problema de inferência de redes um difícil problema de otimização combinatória, para o qual métodos computacionais dispo- níveis falham em relação à exatidão e robustez das predições. Esta tese tem como objetivo investigar o uso de técnicas de ensemble learning como forma de superar as limitações existentes e otimizar o processo de inferência, explorando a diversidade entre um conjunto de modelos. Com este intuito, desenvolvemos métodos computa- cionais tanto para gerar redes diversificadas, como para combinar estas predições em uma solução única (solução ensemble ), e aplicamos esta abordagem a uma série de cenários com diferentes fontes de diversidade a fim de compreender o seu potencial neste contexto específico. Mostramos que as soluções propostas são competitivas com algoritmos tradicionais deste campo de pesquisa e que melhoram nossa capa- cidade de reconstruir com precisão as redes regulatórias genéticas. Os resultados obtidos para a inferência de redes de regulação transcricional e pós-transcricional, duas camadas adjacentes e complementares que compõem a rede de regulação glo- bal, tornam evidente a eficiência e robustez da nossa abordagem, encorajando a consolidação de ensemble learning como uma metodologia promissora para decifrar a estrutura de redes regulatórias genéticas. / In this thesis we are concerned about the reverse engineering of gene regulatory networks from post-genomic data, a major challenge in Bioinformatics research. Gene regulatory networks are intricate biological circuits responsible for govern- ing the expression levels (activity) of genes, thereby playing an important role in the control of many cellular processes, including cell differentiation, cell cycle and metabolism. Unveiling the structure of these networks is crucial to gain a systems- level understanding of organisms development and behavior, and eventually shed light on the mechanisms of diseases caused by the deregulation of these cellular pro- cesses. Due to the increasing availability of high-throughput experimental data and the large dimension and complexity of biological systems, computational methods have been essential tools in enabling this investigation. Nonetheless, their perfor- mance is much deteriorated by important computational and biological challenges posed by the scenario. In particular, the noisy and sparse features of biological data turn the network inference into a challenging combinatorial optimization prob- lem, to which current methods fail in respect to the accuracy and robustness of predictions. This thesis aims at investigating the use of ensemble learning tech- niques as means to overcome current limitations and enhance the inference process by exploiting the diversity among multiple inferred models. To this end, we develop computational methods both to generate diverse network predictions and to combine multiple predictions into an ensemble solution, and apply this approach to a number of scenarios with different sources of diversity in order to understand its potential in this specific context. We show that the proposed solutions are competitive with tra- ditional algorithms in the field and improve our capacity to accurately reconstruct gene regulatory networks. Results obtained for the inference of transcriptional and post-transcriptional regulatory networks, two adjacent and complementary layers of the overall gene regulatory network, evidence the efficiency and robustness of our approach, encouraging the consolidation of ensemble systems as a promising methodology to decipher the structure of gene regulatory networks.

L’établissement de l’identité cellulaire est un phénomène très complexe qui implique pléthore de signaux instructifs intrinsèques et extrinsèques. Cependant, malgré les progrès importants qui ont été faits pour l’identification des régulateurs clés, les liens mécanistiques entre facteurs de transcription, épigénome, et structure de la chromatine lors de la différenciation cellulaire, et de la transformation tumorigénique des cellules, sont peu connus. Pour résoudre ces problématiques nous avons utilisé deux modèles de transition de l’identité cellulaire : la différenciation neuronale et endodermique induites par un même morphogène, l’acide rétinoïque. Concernant la transformation tumorale des cellules nous avons utilisé un système de tumorigenèse par étape de cellules primaires humaines. Nous avons conduit des études intégratives incluant des données transcriptomiques, épigénomiques, et des données concernant l’architecture de la chromatine. Notre approche systématique pour caractériser l’acquisition de l’identité cellulaire, combinée à la modélisation de la transduction du signal, renforce donc nos connaissances sur les mécanismes responsables de la plasticité cellulaire. Une meilleure compréhension des mécanismes régulateurs de l’identité cellulaire non seulement nous éclaire sur les relations de cause à effet entre les différents niveaux de régulation dans la cellule, mais aussi ouvre de nouvelles possibilités en terme de transdifférenciation dirigée. / The cell fate acquisition is a highly complex phenomenon that involves a plethora of intrinsic and extrinsic instructive signals. However, despite the important progress in identification of key regulatory factors of this process, the mechanistic links between transcription factors, epigenome and chromatin structure which coordinate the regulation of cell differentiation and deregulation of gene networks during cell transformation are largely unknown. To address these questions for two model systems of cell fate transitions, namely the neuronal and endodermal cell differentiation induced by the morphogen retinoic acid and the stepwise tumorigenesis of primary human cells, we conducted integrative transcriptome, epigenome and chromatin architecture studies. Through extensive integration with thousands of available genomic data sets, we deciphered the gene regulatory networks of these processes and revealed new insights in the molecular circuitry of cell fate acquisition. The understanding of regulatory mechanisms that underlie the cell fate decision processes not only brings the fundamental understanding of cause-and-consequence relationships inside the cell, but also open the doors to the directed trans-differentiation.

Neuronal nicotinic acetylcholine receptors (nAChRs) are involved in a plethora of fundamental biological processes ranging from muscle contraction to the formation of memories. The studies described in this work focus on the transcriptional regulation of the CHRNB4 gene, which encodes the ß4 subunit of neuronal nAChRs. We previously identified a regulatory sequence (5´– CCACCCCT –3´), or “CA box”, critical for CHRNB4 promoter activity in vitro. Here I report transcription factor interaction at the CA box along with an in vivo analysis of CA box transcriptional activity. My data indicate that Sp1, Sp3, Sox10 and c-Jun interact with the CHRNB4 CA box in the context of native chromatin. Using an in vivo transgenic approach in mice, I demonstrated that a 2.3-kb fragment of the CHRNB4 promoter region, containing the CA box, is capable of directing cell-type specific expression of a reporter gene to many of the brain regions that endogenously express the CHRNB4 gene. Site-directed mutagenesis was used to test the hypothesis that the CA box is critical for CHRNB4 promoter activity in vivo. Transgenic animals were generated in which LacZ expression is driven by a mutant form of the CA box. Reporter gene expression was not detected in any tissue or cell type at ED18.5. Similarly, I observed dramatically reduced reporter gene expression at PD30 when compared to wild type transgenic animals, indicating that the CA box is an important regulatory feature of the CHRNB4 promoter. ChIP analysis of brain tissue from mutant transgenic animals demonstrated that CA box mutation results in decreased interaction of the transcription factor Sp1 with the CHRNB4 promoter. I have also investigated transcription factor interaction at the CHRNB4 promoter CT box, (5´– ACCCTCCCCTCCCCTGTAA –3´) and demonstrated that hnRNP K interacts with the CHRNB4 promoter in an olfactory bulb derived cell line. Surprisingly, siRNA experiments demonstrated that hnRNP K knockdown has no impact on CHRNA5, CHRNA3 or CHRNB4 gene expression. Interestingly, knockdown of the transcription factor Purα results in significant decreases in CHRNA5, CHRNA3 and CHRNB4 mRNA levels. These data indicate that Purα can act to enhance expression of the clustered CHRNA5, CHRNA3 and CHRNB4 genes. Together, these results contribute to a more thorough understanding of the transcriptional regulatory mechanisms underlying expression of the CHRNB4 as well as the CHRNA5 and CHRNA3 genes, critical components of cholinergic signal transduction pathways in the nervous system.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第18701号 / 情博第551号 / 新制||情||97(附属図書館) / 31634 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 阿久津 達也, 教授 山本 章博, 教授 岡部 寿男 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

Yes / Symptoms and signs (symptoms in brief) are the essential clinical manifestations for individualized diagnosis and treatment in traditional Chinese medicine (TCM). To gain insights into the molecular mechanism of symptoms, we develop a computational approach to identify the candidate genes of symptoms. This paper presents a network-based approach for the integrated analysis of multiple phenotype-genotype data sources and the prediction of the prioritizing genes for the associated symptoms. The method first calculates the similarities between symptoms and diseases based on the symptom-disease relationships retrieved from the PubMed bibliographic database. Then the disease-gene associations and protein-protein interactions are utilized to construct a phenotype-genotype network. The PRINCE algorithm is finally used to rank the potential genes for the associated symptoms. The proposed method gets reliable gene rank list with AUC (area under curve) 0.616 in classification. Some novel genes like CALCA, ESR1, and MTHFR were predicted to be associated with headache symptoms, which are not recorded in the benchmark data set, but have been reported in recent published literatures. Our study demonstrated that by integrating phenotype-genotype relationships into a complex network framework it provides an effective approach to identify candidate genes of symptoms. / NSFC Project (61105055, 81230086), China 973 Program (2014CB542903), The National Key Technology R&D Program (2013BAI02B01, 2013BAI13B04), the National S&T Major Special Project on Major New Drug Innovation (2012ZX09503-001-003), and the Fundamental Research Funds for the Central Universities.