Biological data visualization | EPFL Graph Search

Biology data visualization is a branch of bioinformatics concerned with the application of computer graphics, scientific visualization, and information visualization to different areas of the life sciences. This includes visualization of sequences, genomes, alignments, phylogenies, macromolecular structures, systems biology, microscopy, and magnetic resonance imaging data. Software tools used for visualizing biological data range from simple, standalone programs to complex, integrated systems. State-of-the-art and perspectives Today we are experiencing a rapid growth in volume and diversity of biological data, presenting an increasing challenge for biologists. A key step in understanding and learning from these data is visualization. Thus, there has been a corresponding increase in the number and diversity of systems for visualizing biological data. An emerging trend is the blurring of boundaries between the visualization of 3D structures at atomic resolution, visualization

Hybrid Representation and Visualization Methods Applied to Neuronal Microcircuits

Sébastien Lasserre

For the past twenty years, scientific visualization has assisted simulation-based research to explore problems of organized complexity. Environmental research, social sciences, astronomy, and modern physics have exploited the increasing power of computer graphics to interactively visualize three-dimensional models. Of late, neuroscience has witnessed the use of large-scale simulation with biologically detailed neuron models, and aspires to benefit from scientific visualization too. This thesis is the result of a close collaboration with the Blue Brain Project, a neuroscience endeavor working on biologically detailed large-scale simulation of the brain, and tries to answer the following question: How can scientific visualization improve neuroscience insights in the context of simulation-based research? Scientific visualization of a model typically preserves its spatial layout. In continuity with this axiom, we identified in more depth the requirements for a scientific visualization solution in neuroscience. Thus, the preservation of the domain specificity, the hybrid visualization, the validation, and the adaptation to the limitations of the human-visual perception enhance the readability of the generated results. We applied these requirements to elaborate an hybrid representation of neuronal structures, and proposed a unified framework of visualization methods which preserves the domain specificities of neuroscience. This framework was finally assessed in case studies, where neuroscientific investigations about structure, composition, and dynamics of neuronal microcircuit challenge the developed visualization solutions to derive insight. We believe that these representations and visualization methods form a solid ground for the elaboration of virtual instruments dedicated to a new approach of neuroscience, where in silico landscapes of the brain are complementary to the classical in vitro and in vivo investigations.

EPFL2010

Development of an interactive genome browser to visualize and analyse large scale genomic data

Lucas Sinclair

Genomic bioinformatics is a growing and developing field. Indeed, data analysis is becoming an integrative and essential part of any quantitative biological experiment as the technologies evolve and the wet lab methods used generate larger and larger quantities of data. Yet few standards have emerged and a plethora of analytical tools exist, none of which are established as a standard. The difficulties arise early on, even before processing any genomic data, as one first needs to visualize it. Several visualization methods exist, such as the UCSC genome browser, IGB or Argo, but none offer a satisfying interface or set of tools. Stemming from a pre-existing project at the bioinformatics and biostatistics core facility, this study presents a new solution to the multiple difficulties that at present beleaguer the field. A novel genome visualization tool is proposed where the user interface remains simple and incorporates a set of common statistical analysis functions. The software produced, entitled gFeatMiner, is capable of processing large scale genomic datasets for computing descriptive statistics and manipulate them in several ways. The program makes use of modern technologies and infrastructure paving the way for its development into an advanced data mining tool. In the second part of this study, a practical application is worked out. Examining the genes coding for ribosomal proteins in the model organism yeast (Saccharomyces cerevisiae) and using several available sets of data including multiple transcription factor binding profiles in vivo and in vitro, RNA polymerase activity and nucleosome enrichment, we attempt to better understand and reveal cellular mechanisms by clustering the numerous genes together using different criteria and machine learning strategies

2010

Graph Representation Learning with Optimal Transport: Analysis and Applications

Effrosyni Simou

In several machine learning settings, the data of interest are well described by graphs. Examples include data pertaining to transportation networks or social networks. Further, biological data, such as proteins or molecules, lend themselves well to graph-structured descriptions. In such settings, it is desirable to design representation learning algorithms that can take into account the information captured by the underlying structure. This thesis focuses on providing new methods to ingrain geometrical information in end-to-end deep learning architectures for graphs through the use of elements from Optimal Transport theory.First, we consider the setting where the data can be described by a fixed graph. In this setting, each datapoint corresponds to a node of the graph and the edges among nodes signify their pairwise relations. We propose an autoencoder architecture that consists of a linear layer in the encoder and a novel Wasserstein barycentric layer at the decoder. Our proposed barycentric layer takes into account the underlying geometry through the diffusion distance of the graph and provides a way to obtain structure-aware, non-linear interpolations, which lead to directly interpretable node embeddings that are stable to perturbations of the graph structure. Further, the embeddings obtained with our proposed autoencoder achieve competitive or superior results compared to state-of-the-art methods in node classification tasks.Second, we consider the setting where the data consist of multiple graphs. In that setting, the graphs can be of varying size and, therefore, a global pooling operation is needed in order to obtain representations of fixed size and enable end-to-end learning with deep networks. We propose a global pooling operation that optimally preserves the statistical properties of the representations. In order to do so, we take into account the geometry of the representation space and introduce a global pooling layer that minimizes the Wasserstein distance between a graph representation and its pooled counterpart, of fixed size. This is achieved by performing a Wasserstein gradient flow with respect to the pooled representation. Our proposed method demonstrates promising results in the task of graph classification.Overall, in this thesis we provide new methods for incorporating geometrical information in end-to-end deep learning architectures for graph structured data. We believe that our proposals will contribute to the development of algorithms that learn meaningful representations and that take fully into account the geometry of the data under consideration.