Systems-biology based approaches for the identification of genes controlling complex phenotypic traits

下一代测序技术正在彻底改变我们在分子生物学和遗传学方面的研究方式。Illumina是推动下一代技术发展的领先公司之一。DNA测序的成本已经大幅下降，在未来五年内，为许多人进行全基因组测序将成为一项标准技术，Wellcome Trust Sanger Institute的UK10K项目正在进行中。然而，我们才刚刚开始分析和理解这些海量的数据。该项目将为分析下一代序列数据开发新的生物信息学方法。同一物种的个体在DNA序列上存在许多差异，许多变异似乎没有表型效应。但最近的出版物优雅地表明，分析少数个体的蛋白质编码序列足以确定负责单基因性状的基因，例如负责特定遗传遗传疾病的基因（Choi 2009; Ng 2009; Ng 2010）。在这些情况下，单个序列变异的强表型效应允许从候选列表中排除所有先前已知的序列变异。然而，大多数性状不是由单个基因决定的，而是由许多不同的基因决定的。导致这些复杂性状的序列变异将更加难以识别，因为个体变异可能没有任何表型效应，除非它们与其他序列变异结合出现，即我们不能再排除先前已知的序列变异本身。预测单个序列变异对蛋白质产物氨基酸序列的有害影响可以为识别因果变异提供进一步的证据，例如（Ng 2003），尽管这种方法本身不足以识别因果基因。该项目的目的是建立一种利用生物网络与序列分析方法相结合的系统方法，以识别可能对复杂表型性状很重要的序列变异。潜在的假设是，多个序列变异击中涉及功能相关过程的不同蛋白质，将共同导致表型效应。该项目将使用我们从公共数据库和出版物中收集的基因网络、蛋白质网络和代谢网络来检查序列变异对生物系统的功能和潜在影响。这里开发的方法将与一般生物有机体的研究有关；它们也将非常有助于识别导致复杂表型的遗传效应，这可能与植物和动物的育种有关，以及提高我们对克罗恩病、牛皮癣和癌症等复杂疾病的理解。目前，该部门和Illumina正在获取这些疾病患者的大规模序列数据，并将用于分析。该项目的成果将使遗传学、生物信息学、基因和蛋白质网络、系统生物学以及最终疾病过程等领域的研究人员受益。作为该项目的一部分开发的生物信息学软件将作为开源软件免费提供。分子生物学家将受益于我们的软件用户分析他们的序列数据和探索生物网络；因此，该项目将支持设计新的实验方法。崔M.（2009）。科学通报，30(5):996 - 9101。Ng， P. C.（2003）。核酸学报，31(13):3812-3814。吴绍斌（2010）。物理学报，42(1):30-35。吴绍斌（2009）。自然，461(7261):272-276。

Next-generation sequencing technologies are revolutionizing the way we do research in molecular biology and genetics. One of the leading companies driving the development of next-generation technologies is Illumina. The cost of DNA sequencing has dropped so much that within the next five years sequencing whole genomes for many individuals will become a standard technique, as being undertaken in the Wellcome Trust Sanger Institute's UK10K project. However, we are just at the very beginning of analysing and understanding these massive amounts of data. This project will develop new bioinformatics methodologies for the analysis of next-generation sequence data. Individuals of one species show many differences in DNA sequence, many variants appear to be without phenotypic effect. But recent publications demonstrated elegantly that analysing the protein-coding sequences in few individuals is sufficient to identify the gene responsible for monogenic traits, for example responsible for particular genetically inherited diseases (Choi 2009; Ng 2009; Ng 2010). In these cases the strong phenotypic effect of the individual sequence variants allowed to exclude all previously known sequence variants from the candidate lists. However, most traits are not determined by single genes, but rather depend on many different genes. Sequence variants contributing to such complex traits will be much harder to identify, because individual variants might not have any phenotypic effects unless they occur in combination with other sequence variants, i.e. we cannot exclude previously known sequence variants per se any longer. Prediction of the deleterious effects of individual sequence variants on the amino-acid sequence of the protein products can provide further evidence for the identification of causal variants, e.g. (Ng 2003), though this approach on its own is not powerful enough to identify the causal gene(s). The aim of this project is to establish a systems approach utilizing biological networks in combination with sequence analysis methods to identify sequence variants in silico that are likely to be important for complex phenotypic traits. The underlying assumption is that multiple sequence variants that hit different proteins involved in functionally related processes will in combination lead to phenotypic effects. This project will use gene networks, protein networks and metabolic networks that we have collected from public data repositories and publications to examine the function and potential impact of sequence variants on the biological system. The approaches developed here will be relevant for the study of biological organisms in general; they will also be very instrumental for the identification of genetic effects contributing to complex phenotypes, which could be relevant for breeding of plants and animals, as well as to improve our understanding of complex diseases such as Crohn's disease, Psoriasis and Cancer. Large-scale sequence data for individuals suffering from these disorders are currently being obtained within the Department and by Illumina and will be available for analysis. The outcomes of this project will benefit researchers in the areas of genetics, bioinformatics, gene and protein networks, systems biology and ultimately disease processes. Bioinformatics software developed as part of this project will be made available free of charge as open source software. Molecular biologists will benefit as users of our software for the analysis of their sequence data and the exploration biological networks; the project will thus support the design of novel experimental approaches. Choi, M. (2009). Proc Natl Acad Sci U S A 106(45): 19096-19101. Ng, P. C. (2003). Nucleic Acids Res 31(13): 3812-3814. Ng, S. B. (2010). Nat Genet 42(1): 30-35. Ng, S. B. (2009). Nature 461(7261): 272-276.