应数学与统计学院邀请， 美国匹茨堡大学生物信息系Lu Songjian博士近期将访问我校，并做以下学术报告： 

        时间：2015年7月1日上午9:00-11:00
        地点：理科楼407
 

        报告（1）：Cancer Research: Problems, Models, and Algorithms
Usually, in the research of bioinformatics or computational biology, we first formulate a biological problem into a computational problem, such as a statistical problem, a graph problem; we then design an algorithm to solve the computational problem; we finally implement the algorithm and apply it to the data. As the biolog-ical systems are complex, good modules frequently lead to hard computational problems, such as NP-hard problems. Hence, a bioinformatics research project in-cludes not only a good model, but also an efficient algorithm to solve the computa-tional problem in the model.
In this presentation, first, I will introduce the basic knowledge of the molecu-lar biology of the cell, such as the intro-signaling system, the relations between so-matic genome alterations and cancers; then, I will present different mathematical models in cancer research, such as the Bayesian network and the k-path model; next, I will discuss what are NP-hard problems, why the optimal solutions of gen-eral NP-hard problems cannot be found practically, the advantage and disad-vantage of greedy or heuristic algorithms for NP-hard problems; finally, I will in-troduce what are parameterized algorithms and why they can find the optimal so-lutions of the NP-hard problems in our models.
This is a very general talk of the cancer research. The audience will get to know the basic idea and problems of the cancer research in the era of Big Data.
  
        报告（2）：Cancer Research: A signal-based graph models to search cancer driver somatic genome alterations with mutually exclusive property
Somatic genome alterations (SGAs) such as somatic mutations, somatic copy number alterations and epigenomic alterations are major causes of cancers.  In general, SGAs in a tumor can be divided into two types: those that affect cellular signaling proteins, perturb the cellular signaling system, and eventually contribute to cancer initiation and progression are called driver SGAs; and those that do not directly contribute to cancer development are designated as passenger SGAs. A fundamental problem of cancer-genome research is to identify signaling pathways that, when perturbed by driver SGAs, lead to cancer development or affect clinical outcomes for patients. Identification of such pathways will not only advance our understanding of the disease mechanisms underlying cancer, but will also provide guidance for the precision treatment of cancer patients.
It is well known that somatic genome alterations (SGAs) affecting the genes that encode the proteins within a common signaling pathway exhibit mutual exclu-sivity, in which these SGAs usually do not co-occur in a tumor.  With some success, this property has been utilized as an objective function to guide the search for driver mutations within a pathway. However, mutual exclusivity alone is not suffi-cient to indicate that genes affected by such SGAs are in common pathways.  In this talk, I will introduce a novel, signal-oriented framework for identifying driver SGAs.  First, we identify the perturbed cellular signals by mining the gene expression data. Next, we search for a set of SGA events that carries strong information with respect to such perturbed signals while exhibiting mutual exclusivity.   Finally, we design and implement an efficient exact algorithm to solve an NP-hard problem encoun-tered in our approach. Our results indicate that the signal-oriented approach en-hances the ability to find informative sets of driver SGAs that likely constitute sig-naling pathways.
This talk is mainly based on our very recent results accepted in the PLOS Computational Biology, a very top journal in the computational biology. The presen-tation will include detail about research background, model formulation, algorithm designing, results and result evaluation.

        欢迎感兴趣的师生参加！