Keynote Speakers

Prof. M.Luz Calle PROF. M.LUZ CALLE

Biosciences Department, University of Vic - Central University of Catalonia, Spain

Challenges of Microbiome Compositional Data Analysis

Understanding human-microbiome relationship and how it can be modulated is a frontier for preventive medicine and for the medical management of chronic diseases and represents an opportunity for food and pharma industry. Although the human microbiome has long been known to influence human health and disease, until recently, the composition and properties of the human microbiota were largely unknown because their study was limited to in vitro studies where specific microorganisms were isolated and cultured. High-throughput DNA sequencing technologies have revolutionized this field by allowing the study of the genomes of all microorganisms of a given environment. From a statistical point of view, microbiome data analysis is challenging because it involves high-dimensional structured multivariate sparse data. Furthermore, microbiome data is compositional, since raw abundances and the total number of sequences for each sample are not by itself informative, as they depend on technical issues such as laboratory sample preparation and sequencing depth. In this talk I will discuss the most important statistical challenges of microbiome data analysis, in particular, the effects of ignoring the compositional structure of microbiome data and will present new procedures for the analysis of microbiome compositional data.

Dr. Ana Cvejic DR. ANA CVEJIC

Wellcome Trust Sanger Institute, Cambridge, UK

Single-Cell RNA-Sequencing Provides Insights Into Fate Decisions in Haematopoiesis

Mammalian blood formation is the most intensely studied system of stem cell biology, with the ultimate aim to obtain a comprehensive understanding of the molecular mechanisms controlling fate-determining events. A single cell type, the haematopoietic stem cell (HSC), is responsible for generating more than 10 different blood cell types throughout the lifetime of an organism. This diversity in the lineage output of HSCs is traditionally presented as a stepwise progression of distinct, transcriptionally homogeneous populations of cells along a hierarchical differentiation tree. However, most of the data used to explain the molecular basis of lineage differentiation and commitment were derived from populations of cells isolated based on well-defined cell surface markers. An inherent problem with this approach is that the presence of specific cell surface markers does not directly reflect the transcriptional state of a cell.
Here we used a marker-free approach to computationally reconstruct the blood lineage tree in zebrafish and order cells along their differentiation trajectory, based on their global transcriptional differences. By reconstructing their developmental chronology computationally, we were able to place each cell along a continuum from stem cell to mature cell, refining the traditional lineage tree. Within the population of transcriptionally similar stem and progenitor cells our analysis revealed considerable cell-to-cell differences in their probability to transition to another, committed state. This suggested that although global transcriptional changes before and after the branching point were continuous, the probability of a cell progressing to any of the committed states was determined only by a subset of highly relevant genes. Therefore, cells that were transcriptionally similar overall could have a high probability of differentiation to distinct cell types. Once the cell fate decision was executed, the progression of cells along the continuum is characterised by a highly coordinated transcriptional program, displaying simultaneous suppression of genes involved in cell proliferation and ribosomal biogenesis and increased expression of lineage specific genes. Our comparative analysis between zebrafish, mouse and human across seven different haematopoietic cell types, including innate lymphocytes (ILCs), revealed a high level of conservation of blood cell type specific genes. The lowest conservation was observed for lymphocytes, possibly reflecting their adaptation to fish specific pathogens and virulence factors.

Prof. Uzay Kaymak PROF. UZAY KAYMAK

School of Industrial Engineering, Eindhoven University of Technology, The Netherlands

Paths to transparency for computational intelligence models in clinical decision support

As electronic medical records become the norm for documenting medical history of patients, reuse of the data generated during the care process opens new ways of supporting clinical decisions. Advanced data analysis techniques, such as computational intelligence models that make secondary use of medical data are accepted more and more in clinical applications. In these applications, care professionals find it important to have transparent models whose behavior can be understood well. In this presentation, we discuss what transparency might encompass in clinical decision support and how various computational intelligence models relate to interpretability and transparency. We discuss different examples that use fuzzy set theory to develop transparent models for supporting clinical decisions.

Invited speakers

Prof. Marco Masseroli PROF. MARCO MASSEROLI

Politecnico di Milano, Italy

Invited to the special session "Machine Learning in Healthcare Informatics and Medical Biology"

Processing of big heterogeneous genomic datasets for tertiary analysis of Next Generation Sequencing data

We previously proposed a paradigm shift in genomic data management, based on the Genomic Data Model (GDM) for mediating existing data formats and on the GenoMetric Query Language (GMQL) for supporting, at a high level of abstraction, data extraction and the most common data-driven computations required by tertiary data analysis of Next Generation Sequencing datasets. Here, we present a new GMQL-based system with enhanced accessibility, portability, scalability and performance.
The new system has a well-designed modular architecture featuring: (i) an intermediate representation supporting many different implementations (including Spark, Flink and SciDB); (ii) a high-level technology-independent repository abstraction, supporting different repository technologies (e.g., local file system, Hadoop File System, database or others); (iii) several system interfaces, including a user-friendly Web-based interface, a Web Service interface, and a programmatic interface for Python language. Biological use case examples, using public ENCODE, Roadmap Epigenomics and TCGA datasets, demonstrate the relevance of our work


For any request, please contact

Programme co-chairs

Paolo Cazzaniga [Web] -- Department of Human and Social Sciences, University of Bergamo, Italy
Ivan Merelli [Web] -- Institute of Biomedical Technologies, National Research Council, Italy
Daniela Besozzi [Web] -- Department of Informatics, Systems and Communication, University of Milano-Bicocca, Italy

Supported by

CNR    unibg    DISCo    DISCo