科研动态
学术讲座
现在位置: 首页  科研动态  学术讲座
  • 计算系统生物学讨论会
  • 浏览量:10 发布人: 发布时间:2019-03-12

  •            会议日程         2019316

    (地点:精正楼二楼学术报告厅)

    时间

    报告人

    题目

    主持人

    08:5009:00

    开幕式

    ²苏州大学数学科学学院领导致辞


    09:009:30

    周天寿

    非马氏反应系统的随机动力学分析

    李铁军

    9:30-10:00

    靳 祯

    网络上传播动力学几点思考

    10:0010:15

    茶歇

    10:1510:45

    陈洛南

    Cell-specific Network Constructed by Single-cell RNA Sequencing Data

    周天寿

    10:4511:15

    雷锦誌

    基于单细胞RNA-seq 的干性指标与干细胞增殖的数学模型

    11:15-11:45

    王勇

    Seeking Cell-Type-Indicative Marker from single cell RNA-seq data by optimization model optimization model

    12:0013:30

    午餐

    13:3014:00

    林伟

    TBD

    陈洛南

    14:0014:30

    李铁军

    Fused D-Trace Loss for differential network inference

    14301500

    张磊

    Phase field modeling of membrane fusion

    15:0015:15

    茶歇

    15:1515:45

    吴凌云

    Sample-specific differentially expressed gene analysis for heterogeneous data using partial order graph

    林伟

    15:4516:15

    张世华

    TBD

    16:15-16:45

    刘锐

    Single-sample landscape entropy reveals the imminent phase transition during disease progression

    16:45-17:15

    周栋焯

    Chaotic or nonchaotic dynamics in neuronal networks?

    报告摘要

    非马氏反应系统的随机动力学分析

    周天寿(中山大学)

    摘要:对于反应系统,传统的研究假设反应事件是马氏的。然而,由于反应的非完全混合或由于一个宏观分子的产生需要经过若干小的反应步,因此反应通常是以非马氏的方式发生。我将介绍研究一般(即具有任意等待时间分布)反应网络的建模与分析方法。

      

      

      

    网络上传播动力学几点思考

    靳 祯

    摘要:传染病的传播动力学研究已有大量的成果,从数学工具看,有确定性的非空间结构的ODE 模型,也有确定性的连续空间结构PDE 模型,以及离散空间的网络模型,还有复杂网络上的传播模型,以及随机意义下的随机动力学模型。本报告将聚焦不同模型的相互转化及关联性,以及空间网络结构及传播的时间尺度,网络上的聚集性、相关性以及网络的路障对传播的影响。

      

      

      

      

    Cell-specific Network Constructed by Single-cell RNA Sequencing Data

    陈洛南(中科院上海生命科学院)

     Abstract: Single-cell RNA sequencing (scRNA-seq) is able to give an insight into the gene-gene interactions or transcriptional networks among cell populations based on the sequencing of a large number of cells. However, traditional network methods are limited to the grouped cells instead of each single cell, and thus the heterogeneity of single cells will be erased. We present a new method to construct a cell-specific network (CSN) for each single cell from scRNA-seq data (i.e. one network for one cell), which transforms the data from “unstable” gene expression form to “stable” gene association form on a single-cell basis. In particular, it is for the first time that we can identify the gene interactions/network at a single-cell resolution level. By CSN method, scRNA-seq data can be analyzed for clustering and pseudo-trajectory from network perspective by any existing method, which opens a new way to scRNA-seq data analyses. In addition, CSN is able to find key genes from a network viewpoint, find “dark” genes that have no significant difference in gene expression level, but in network degree level, and find new cell types. Experiments on various scRNA-seq datasets demonstrated the improvement of CSN over existing methods in terms of accuracy and robustness. 

      

      

      

      

    基于单细胞RNA-seq 的干性指标与干细胞增殖的数学模型

    雷锦誌(清华大学)

    摘要本报告将介绍两部分内容:(1)基于单细胞RNA-seq 的干性指标的建立;(2)基于干性差异的异质性干细胞增殖模型及其在癌症演变动力学中的应用。我还将探讨相关问题未来可能的研究方向。

      

      

    Seeking Cell-Type-Indicative Marker from single cell RNA-seq data by optimization model

    王勇(中科院系统所)

     Single cell RNA-seq data offers us new resource and resolution to study cell type identity and its conversion. This task is challenging in dealing with noise, sparsity, and poor annotation for single cells. By detecting cell-type-indicative markers, it is promising to help denoising, clustering and cell type annotation. Here, we develop an optimization model, scTIM, to detect cell-type-indicative markers. scTIM is based on a multi-objective optimization framework to simultaneously maximize gene specificity by considering gene-cell relationship, maximize gene’s ability to reconstruct cell-cell relationships, and minimize gene redundancy by considering gene-gene relationships. Furthermore, we propose a consensus optimization framework for robust solution. Experimental results on three diverse single cell RNA-seq datasets show that the revealed cell-type-indicative markers have advantages in improving the clustering, revealing the biological functions, and reconstructing cell development trajectory. Application to the large scale mouse cell atlas data identifies critical markers for 15 tissues.

      

      

    Fused D-Trace Loss for differential network inference

    李铁军(北京大学)

    Abstract: The correlation or connection of two genes in biological networks are of great importance in biological and medical research. In this talk, I will introduce a new loss function called fused D-trace loss to estimate the differential matrices. A very efficient algorithm was developed based on alternating and coordinate descent methods. The convergence and consistency of the algorithm is also established. Our method outperforms existing methods in both simulation examples and real data.

      

      

    Phase field modeling of membrane fusion

    张磊(北京大学)

    Abstract: Membrane fusion is involved in many cellular processes such as endocytosis and exocytosis, synaptic release, etc. Many efforts have been made in studying membrane fusion process, however, the application of atomistic model is greatly restricted by the short time interval and limited length scale and the bilayer structure has not been studied in continnum model. We developed phase field model containing both short-range and long-range free energy derived from particle model to study the process of membrane fusion. The transition path from two separate membranes to a fusion pore is constructed using a combination of Cahn-Hilliard equation and string method. With solvent between two apposed membrane, a leaky transition path is derived through an asymmetrical configuration stalk-hole complex. Without solvent, the classical symmetry transition pathway and the pathway through IMI(inverted micellar intermediate) configuration are found. The joint work with Anchang Shi (McMaster).

      

      

    Sample-specific differentially expressed gene analysis for heterogeneous data using partial order graph

    吴凌云(中科院系统所)

    Abstract: High-throughput experimental techniques such as microarray and RNA-Seq have created an unprecedented opportunity for characterize vary samples of different conditions in gene level. One important fundamental task in high-throughput gene expression analysis is to identify differentially expressed gene (DEG) between different conditions, e.g. disease and normal samples. The high heterogeneity in individual samples brings great challenge to this task. We proposed a method called netDEG for identifying differentially expressed genes of individual samples with high heterogeneity. By detecting the variation of gene pair expression ratios, a partial order graph is constructed for each sample, in which the differentially expressed genes tend to have high degrees. The significant up- and down- regulated genes are further identified through a statistical model. Our method shows great superiority over other widely applied DEG identification methods, including RNA-seq oriented and microarray oriented, on both simulated and real dataset. The superior performance of our method is further evaluated on a wide range of real public TCGA and GEO datasets which cover a variety of cancer types. The method has been implemented in the R package Corbi available in CRAN (https://cran.r-project.org/package=Corbi).

      

      

      

    Single-sample landscape entropy reveals the imminent phase transition during disease progression

    刘锐(华南理工大学)

    Abstract: The time evolution or dynamic change of many biological systems during disease progression is not always smooth but occasionally abrupt, that is, there is a tipping point during such a process at which the system state shifts from the normal state to a disease state. It is challenging to predict such disease state with the measured omics data, in particular when only a single sample is available.  In this study, we developed a novel approach, i.e., single-sample landscape entropy (SLE) method, to identify the tipping point during disease progression with only one sample data. Specifically, by evaluating the disorder of a network projected from a single-sample data, SLE effectively characterizes the criticality of this single sample network in terms of network entropy, thereby capturing not only the signals of the impending transition but also its leading network, i.e. dynamic network biomarkers. Using this method, we can characterize sample-specific state during disease progression and thus achieve the disease prediction of each individual by only one sample.

      

      

    Chaotic or nonchaotic dynamics in neuronal networks?

    周栋焯

    Abstract: In this talk, I will address two issues related to the dynamical stability of integrate-and-fire type neuronal networks. i) Whether there exists chaotic dynamics in I&F neuronal networks. It has been shown that a single integrate-and-fire (I&F) neuron under a general time-dependent stimulus cannot possess chaotic dynamics despite the firing-reset discontinuity. However, whether the network dynamics can be chaotic was an open question. Through correct renormalization and augmented dynamics, we extend the classical Lyapunov exponents (LEs) theory, which is established for smooth dynamical systems, to the I&F like network dynamics and provide a stable and accurate numerical algorithm to compute the LEs of these non-smooth dynamical systems. ii) Whether the irregular firing activity of balanced neuronal networks arise from chaotic dynamics. Some previous studies have shown that chaotic dynamics in the balanced state, i.e., one with balanced excitatory and inhibitory inputs into cortical neurons, is the underlying mechanism for the irregularity of neural activity. However, we show that the balanced state robustly persists in current-based integrate-and-fire neuronal networks with delta-pulse coupling within a broad range of parameters and mathematically prove that the largest Lyapunov exponent of this type of neuronal networks is always negative. Therefore, the irregularity of balanced neuronal networks need not arise from chaos.

      

      

      

      

      

      

      

      

      

      

      


苏州大学数学科学学院 Copyright 2016.All Rights Reserved. 管理后台
电话:0512-65112637 传真:0512-65112637  E-mail:sxxy@suda.edu.cn  地址:苏州市十梓街1号 邮编:215006 苏ICP备06032411号