Intelligent systems in biology: why the excitement?

Biology has rapidly become a data-rich, information-hungry science because of recent massive data generation technologies. Our biological colleagues are designing more clever and informative experiments because of recent advances in molecular science. These experiments and data hold the key to the deepest secrets of biology and medicine, but we cannot fully analyze this data due to the wealth and complexity of the information available. The result is a great need for intelligent systems in biology. There are many opportunities for intelligent systems to help produce knowledge in biology and medicine. Intelligent systems probably helped design the last drug your doctor prescribed, and they were probably involved in some aspect of the last medical care you received. Intelligent computational analysis of the human genome will drive medicine for at least the next half-century. Intelligent systems are working on gene expression data to help understand genetic regulation and ultimately the regulated control of all life processes including cancer, regeneration, and aging. Knowledge bases of metabolic pathways and other biological networks make inferences in systems biology that, for example, let a pharmaceutical program target a pathogen pathway that does not exist in humans, resulting in fewer side effects to patients. Modern intelligent analysis of biological sequences produces the most accurate picture of evolution ever achieved. Knowledge-based empirical approaches currently are the most successful method known for general protein structure prediction. Intelligent literature-access systems exploit a knowledge flow exceeding half a million biomedical articles per year. Machine learning systems exploit heterogenous online databases whose exponential growth mimics Moore's law.     

Perspectives on systems biology

Systems biology is a new field in biology that aims at system-level understanding of biological systems, such as cells and organisms. Molecular biology has already made remarkable contribution to our understanding of biological systems, and its current focus is on the identification of genes and the functions of their products; that is, on the components of systems. There is no doubt that molecular biology will progress even faster and finally identify all the components of biological systems. As such a moment approaches, major importance need to be placed on the establishment of methodologies and techniques that enable us to understand biological systems as systems. This paper overviews the field of systems biology. Hiroaki Kitano, Ph.D.: Hiroaki Kitano is a Senior Researcher at Sony Computer Science Laboratories, Inc., a Project Director of Kitano Symbiotic Systems Project, Japan Science and Technology Corporation and a visiting associate at California Institute of Technology. He received a B.A. in Physics from the International Christian University, Tokyo, and a Ph.D. in Computer Science from Kyoto University. Since 1988, he has been a visiting researcher at the Center for Machine Translation at Carnegie Mellon University. Kitano received Computers and Thought Award from the International Joint Conferences on Artificial Intelligence in 1993. His research interests include RoboCup, computational molecular biology, engineering use of the mophogenesis process, and evolutionary systems.     

Accomplishments and challenges in bioinformatics

A new IT discipline - bioinformatics - fuses computing, mathematics and biology to meet the many computational challenges in modern molecular biology and medical research. The two major themes in bioinformatics - data management and knowledge discovery - rely on effectively adopting techniques developed in IT for biological data, with IT scientists playing an essential role. The future of molecular biology and biomedicine will greatly depend on advances in informatics. As we review researchers' many achievements in bioinformatics, we're confident that the marriage between molecular biology and information technology is a happy one. Accomplishments in bioinformatics have advanced molecular biology and information technology. Although many computational challenges lie ahead, more fruitful outcomes of this successful multidisciplinary marriage are likely.     

Some Algorithmic Challenges in Genome-Wide Ortholog Assignment

Genome-scale assignment of orthologous genes is a fundamental and challenging problem in computational biology and has a wide range of applications in comparative genomics, functional genomics, and systems biology. Many methods based on sequence similarity, phylogenetic analysis, chromosomal syntenic information, and genome rearrangement have been proposed in recent years for ortholog assignment. Although these methods produce results that largely agree with each other, their results may still contain signi...     

Spectral distortion measures for biological sequence comparisons and database searching

In bioinformatics and computational biology, methods for biological sequence comparison play the most important role for the interpretation of complex nucleotide and protein data such as the inference of relationships between genes, proteins and species; and the discovery of novel protein structures and functions. This type of inference is derived by sequence similarity matching on the databases of biological sequences. As many entire genomes have being determined at a rapid rate, computational methods for comparing genomic and protein sequences will be more essential for probing the complexity of genes, genomes, and molecular machines. In this paper we introduce a pattern-comparison algorithm, which is based on the mathematical concepts of linear predictive coding and its cepstral-distortion measures for the analyses of both DNA and protein sequences. The results obtained from several experiments on real datasets have shown the effectiveness of the proposed approach.     

Promoter analysis of co-regulated genes in the yeast genome.

The use of high density DNA arrays to monitor gene expression at a genome-wide scale constitutes a fundamental advance in biology. In particular, the expression pattern of all genes in Saccharomyces cerevisiae can be interrogated using microarray analysis where cDNAs are hybridized to an array of more than 6000 genes in the yeast genome. In an effort to build a comprehensive Yeast Promoter Database and to develop new computational methods for mapping upstream regulatory elements, we started recently in an on going collaboration with experimental biologists on analysis of large-scale expression data. It is well known that complex gene expression patterns result from dynamic interacting networks of genes in the genetic regulatory circuitry. Hierarchical and modular organization of regulatory DNA sequence elements are important information for our understanding of combinatorial control of gene expression. As a bioinformatics attempt in this new direction, we have done some computational exploration of various initial experimental data. We will use cell-cycle regulated gene expression as a specific example to demonstrate how one may extract promoter information computationally from such genome-wide screening. Full report of the experiments and of the complete analysis will be published elsewhere when all the experiments are to be finished later in this year (Spellman, P.T., et al. 1998. Mol. Biol. Cell 9, 3273-3297).     

A clinical molecular scanner: the Melanie project

We developed an expert system to analyze and interpret protein maps. This system, Melanie (medical electrophoresis analysis interactive expert), can distinguish between normal and cirrhotic liver and identify various types of cancer on the basis of protein patterns in biopsy specimens. Our findings suggest that some diseases associated with toxic compounds or modifications of the human genome can be diagnosed by expert systems that analyze protein maps. The combination of protein mapping and computer analysis could result in a clinically useful "molecular scanner". The massive amount of information analyzed and stored in such studies requires new strategies, including centralized databases and image transmission over networks. Increased understanding of protein expression and regulation will enhance the importance of the human genome project in medicine and biology.     

Langevin dynamics simulations of micromechanics on graphics processors

Due to the very long timescales involved (ms-s), theoretical modeling of fundamental biological processes including folding, misfolding, and mechanical unraveling of biomolecules, under physiologically relevant conditions, is challenging even for distributed computing systems. Graphics Processing Units (GPUs) are emerging as an alternative programming platform to the more traditional CPUs as they provide high raw computational power that can be utilized in a wide range of scientific applications. Using a coarse-grained Self Organized Polymer (SOP) model, we have developed and tested the GPU-based implementation of Langevin simulations for proteins (SOP-GPU program). Simultaneous calculation of forces for all particles is implemented using either the particle based or the interacting pair based parallelization, which leads to a ∼90-fold acceleration compared to an optimized CPU version of the program. We assess the computational performance of an end-to-end application of the SOP-GPU program, where all steps of the algorithm are running on the GPU, by profiling the associated simulation time and memory usage for a number of small proteins, long protein fibers, and large-size protein assemblies. The SOP-GPU package can now be used in the theoretical exploration of the mechanical properties of large-size protein systems to generate the force-extension and force-indentation profiles under the experimental conditions of force application, and to relate the results of singlemolecule experiments in vitro and in silico.     

基因组重组排序算法综述

随着快速测序技术的发展,对大规模DNA分子的研究与其中的基因相对次序有关。基因组重组是计算生物学的一个重要研究领域,是基因组在基因水平比较分析的基础。其研究目标是找最短的重组操作序列,将一种基因组转变为另一种基因组。基于分子生物学的实验证明,这种序列有助于估计不同基因组间的进化事件。基因组进化过程虽然非常复杂,但可用3种基本的重组操作模拟,即反转（reversal）、移位（transloeation）和转位（transposition）。本文讨论了这些操作相关的重组算法以及各种排序距离的计算方法。     

Recent trends in knowledge and data integration for the life sciences

Abstract: The bioscience field has seen some spectacular advances in genomic and proteomic technologies that are able to deliver vast quantities of information on cellular activity. Such technologies are of critical importance to biology, medical science and in drug discovery. However, living systems are highly complex and to fully exploit these technologies requires knowledge at many different levels. Information such as genome sequence data, gene expression data, protein-to-protein interactions and metabolic pathways is required to understand the complexity of biological processes. The challenge for bioinformatics is to tackle the problem of fragmentation of knowledge by integrating the many sources of heterogeneous information into a coherent entity. Another problem is that the high level of biological complexity and the fragmented nature of biological research has meant that it is difficult to keep fully conversant with the latest research and discoveries. Progress in one area of biology may have implications for other areas but the dissemination of this knowledge is not straightforward; difficulties such as differences in naming conventions for genes and biological processes has led to confusion and the lack of productivity. This paper reviews the most recent research to overcome the fragmentation problem where technologies such as text mining and ontologies are used within the knowledge discovery process and the specific technical challenges they address.     

Clustering of high throughput gene expression data

High throughput biological data need to be processed, analyzed, and interpreted to address problems in life sciences. Bioinformatics, computational biology, and systems biology deal with biological problems using computational methods. Clustering is one of the methods used to gain insight into biological processes, particularly at the genomics level. Clearly, clustering can be used in many areas of biological data analysis. However, this paper presents a review of the current clustering algorithms designed especially for analyzing gene expression data. It is also intended to introduce one of the main problems in bioinformatics – clustering gene expression data – to the operations research community.     

Putting more genetics into genetic algorithms

The majority of current genetic algorithms (GAs), while inspired by natural evolutionary systems, are seldom viewed as biologically plausible models. This is not a criticism of GAs, but rather a reflection of choices made regarding the level of abstraction at which biological mechanisms are modeled, and a reflection of the more engineering-oriented goals of the evolutionary computation community. Understanding better and reducing this gap between GAs and genetics has been a central issue in an interdisciplinary project whose goal is to build GA-based computational models of viral evolution. The result is a system called Virtual Virus (VIV). VIV incorporates a number of more biologically plausible mechanisms, including a more flexible genotype-to-phenotype mapping. In VIV the genes are independent of position, and genomes can vary in length and may contain noncoding regions, as well as duplicative or competing genes. Initial computational studies with VIV have already revealed several emergent phenomena of both biological and computational interest. In the absence of any penalty based on genome length, VIV develops individuals with long genomes and also performs more poorly (from a problem-solving viewpoint) than when a length penalty is used. With a fixed linear length penalty, genome length tends to increase dramatically in the early phases of evolution and then decrease to a level based on the mutation rate. The plateau genome length (i.e., the average length of individuals in the final population) generally increases in response to an increase in the base mutation rate. When VIV converges, there tend to be many copies of good alternative genes within the individuals. We observed many instances of switching between active and inactive genes during the entire evolutionary process. These observations support the conclusion that noncoding regions serve as scratch space in which VIV can explore alternative gene values. These results represent a positive step in understanding how GAs might exploit more of the power and flexibility of biological evolution while simultaneously providing better tools for understanding evolving biological systems.     

Computational Challenges in Deciphering Genomic Structures of Bacteria

This article addresses how the functionalities of the cellular machinery of a bacterium might have constrained the genomic arrangement of its genes during evolution and how we can study such problems using computational approaches, taking full advantage of the rapidly increasing pool of the sequenced bacterial genomes, potentially leading to a much improved understanding of why a bacterial genome is organized in the way it is. This article discusses a number of challenging computational problems in elucidat...     

Multiple-kernel SVM based multiple-task oriented data mining system for gene expression data analysis

Gene expression profiling using DNA microarray technique has been shown as a promising tool to improve the diagnosis and treatment of cancer. Recently, many computational methods have been used to discover maker genes, make class prediction and class discovery based on gene expression data of cancer tissue. However, those techniques fall short on some critical areas. These included (a) interpretation of the solution and extracted knowledge. (b) Integrating various sources data and incorporating the prior knowledge into the system. (c) Giving a global understanding of biological complex systems by a complete knowledge discovery framework. This paper proposes a multiple-kernel SVM based data mining system. Multiple tasks, including feature selection, data fusion, class prediction, decision rule extraction, associated rule extraction and subclass discovery, are incorporated in an integrated framework. ALL-AML Leukemia dataset is used to demonstrate the performance of this system.     

Classification of proteins multiple-labelled and single-labelled with protein functional classes

Advances in high-throughput genome sequencing technology have led to an explosion in the amount of sequence data that are available. The determination of protein function using experimental techniques is time-consuming and expensive; the use of machine-learning techniques rapidly to assess protein function may be useful in streamlining this process. The problem of assigning functional classes to proteins is complicated by the fact that a single protein can participate in several different pathways and thus can have multiple functions. We have developed a tree-based classifier that is capable of handling multiple-labelled data and gaining an insight into the multi-functional nature of proteins. We call the resulting tree a recursive maximum contrast tree (RMCT) and the resulting classifier a multiple-labelled instance classifier (MLIC). We investigate the synergy of machine-learning-based ensemble methods and physiochemical-based feature augments. We test our algorithm on protein phylogenetic profiles generated from 60 completely sequenced genomes and we compare our results with those achieved by algorithms such as support vector machines and decision trees.     

Motion Understanding: Task-Directed Attention and Representations that Link Perception with Action

This short paper outlines my position on a future direction for computational research on visual motion understanding. The direction combines motion perception, visual attention, action representation and computational vision. Due the breadth of literature in these areas, the paper cannot present a comprehensive review of any one topic. The review is a selective one, a selection that attempts to make some particular points. I claim that task-directed attentive processing is a largely unexplored dimension in the computational motion field. I recount in the context of motion understanding a past argument that in order to make vision systems general, attention is one of the components of the strategy. No matter how sophisticated the methods become for extracting motion information from image sequences, it will not be possible to achieve the goal of human-like performance without integrating the optimization of processing that attention provides. Virtually all past surveys of computational models of motion processing completely ignore attention. However, the concept has crept into work over the years in a variety of ways. A second claim is that the biology of attention offers some interesting insights to guide future development. Many computational authors had previously commented that too little is known about how biological vision systems use task-directed attention in motion processing; this is no longer true. Here, I briefly summarize biological evidence that attentive processing affects all aspects of visual perception including motion, and again emphasize that this paper does not do justice to the breadth and depth of the field. New findings provide a critical link between the perception of visual actions and their execution. Together these findings point to a strategy for motion understanding closely related to that presented more than two decades ago.     

Proteomic identification of proteins in the human brain: Towards a more comprehensive understanding of neurodegenerative disease

Proteomics has revealed itself as a powerful tool in the identification and determination of proteins and their biological significance. More recently, several groups have taken advantage of the high-throughput nature of proteomics in order to gain a more in-depth understanding of the human brain. In turn, this information has provided researchers with invaluable insight into the potential pathways and mechanisms involved in the pathogenesis of several neurodegenerative disorders, e.g., Alzheimer and Parkinson disease. Furthermore, these findings likely will improve methods to diagnose disease and monitor disease progression as well as generate novel targets for therapeutic intervention. Despite these advances, comprehensive understanding of the human brain proteome remains challenging, and requires development of improved sample enrichment, better instrumentation, and innovative analytic techniques. In this review, we will focus on the most recent progress related to identification of proteins in the human brain under normal as well as pathological conditions, mainly Alzheimer and Parkinson disease, their potential application in biomarker discovery, and discuss current advances in protein identification aimed at providing a more comprehensive understanding of the brain.