Algorithms and Methods for small RNA studies

下一代测序技术的出现提供了数十亿个从短序列到全基因组的DNA元素。事实上，天文数据已经并将从全基因组中产生，从基因组中表达的基因对实验、环境、地理空间和/或与疾病相关的条件的影响。快速而系统的功能分类，以及识别这些序列的生物学作用，构成了生物信息学和计算生物学的一个主要问题。我的发现研究涉及从组学数据预测小rna的功能分类和进化模式的方法设计。这种预测方法依赖于两个主要特点：比较组学方法的强大功能和将组学数据与几个相关的非组学数据相结合的综合方法。利用文献中存在的深度基因组知识和大量异构产生的数据，我计划将这些知识表示出来，使得出结论的推理自动化，学习机器适应新的特征，如祖先序列和隐藏模式，并识别新的小RNA模式。这项发现资助将允许利用人工智能、机器学习、图形模型、比较基因组学和组学研究的综合方法的创新方法。最终目标包括提供生物信息学工具，以高精度预测和分类这些小RNA类的功能和调控机制。我的发现计划将集中在三个主要目标上：1)设计准确的方法来识别不同类别的srna；2)开发从RNAseq数据中挖掘隐藏模式和特征的方法，以改进分类；3)设计利用比较基因组学和sRNA分类进化起源的方法来改进分类。该提案将提供新的备受期待的工具，包括在主要组学平台构建的模块中。这一发现计划的结果将使我们更好地了解影响超过60%的蛋白质转录基因的sRNA生物学功能；提高我们对途径调控机制和分子标记开发的理解，帮助植物育种者选择更好的、更耐生物和非生物胁迫的作物；开发sRNA肿瘤生物标志物；针对几种涉及sRNAs调控基因的疾病，如心脏病、炎症和自身免疫性疾病，开发不同的治疗方法。由于其巨大的经济价值，SRNAs将继续成为未来十年医药农业行业的主要战略目标。这项发现计划将提供生物信息学、人工智能、分子生物学方面的高素质人才培训，并在高影响力的开放获取期刊上发表多篇论文。

The emergence of Next Generation Sequencing provides billions of DNA elements from short sequence to full genomes. In fact, astronomic data have been and will be produced from whole genome, expressed genes from genomes regarding to their implication to either experimental, environmental, geospatial, and/or disease related conditions. The rapid and systematic functional classification, as well as the identification of the biological roles of such sequences, constitutes a major issue in bioinformatics and computational biology. My discovery research concerns the design of methods to predict from omics data the functional classification as well as evolutionary patterns of small RNAs. Such prediction methods rely on two main characteristics: the powerfulness of comparative omics methods and integrative approaches combining omics data to several related non-omics data. With the deep genomic knowledge present in the literature and the massive heterogeneous produced data, I plan to represent the knowledge, to automate the reasoning for drawing conclusions, to learn machines to adapt at new characteristics such as ancestral sequences and hidden patterns, and to recognize novel small RNA patterns. This discovery grant will permit innovative methods exploiting artificial intelligence, machine learning, graphical models, comparative genomics and integrative approaches for omics studies. The ultimate goal consisting of providing bioinformatics tools that predict and classify with high accuracy the function and regulatory mechanisms of these small RNA classes. My discovery program will focus on three main objectives: 1) design accurate methods for the identification of different classes of sRNAs; 2) develop method exploring hidden patterns and features from RNAseq data to improve the classification; 3) design methods exploiting comparative genomics and evolutionary origin of sRNA classes to improve classification. This proposal will provide new highly expected tools to be included in module constructed for main omics platforms. The outcome of this discovery program will allow a better understanding of sRNA biological functions that affect more than 60% of protein transcribed genes; improve our understanding of pathway regulation mechanisms and molecular marker development that help plant breeder to select better crops with greater tolerance to biotic and abiotic stresses; develop sRNA cancer biomarker; develop different therapy for several diseases that involved genes regulated by sRNAs such as Cardiac, inflammatory and autoimmune diseases. SRNAs will remain a main strategic target for pharmaceutical agricultural industries for this decade due to its huge economical values. This discovery proposal will provide training of high-qualified personals in bioinformatics, artificial intelligence, molecular biology as well as several papers in high impact open access journals.