RegaDB AREVIR Meeting, Bonn 10-11 April 2008 Pieter Libin Agenda - - PowerPoint PPT Presentation

RegaDB

AREVIR Meeting, Bonn 10-11 April 2008 Pieter Libin

Agenda

RegaDB software ViroLab data integration ViroLab – EUResist collaboration

Introduction

RegaDB

− HIV Data and Analysis Management environment − To be used by

Clinicians Researchers

Two main usage scenarios:

− Runs locally

On a server in your institute On your laptop

− On a collaboration web site − Easy access with Web browser

Objective

A free and open-source database system Facilitate research collaborations

− Easily interchange data, keeping track of data origin − Create collaborative databases keeping track of dataset versions − Easily publish data sets to Stanford HIVDB

Empower clinicians with analysis tools

− Patient history graph − Sequence analyses

HIV typing and subtyping Genotypic analyses for resistance interpretation Third party analyses as “web services”

− Resistance reports

Data Entities

Patient Therapy Dataset Test Viral Isolate Attribute

Attributes

Annotate Patient with information

− Adapts to “your (research) interest” − Typically clinical and epidemiological information

Examples

− Gender − HIV transmission risk group − Most of information collected in SPREAD

Attributes may be grouped (like “SPREAD”)‏

Viral Isolates

Viral Isolates Insertion Nucleotide Sequence Amino Acid Sequence Mutation Protein Information automatically inferred

Treatment

Consists of a drug combination

− commercial or generic representation − dosage information

Start & end date

− support for therapy ending motivation

Commercial drug representation can be

automatically mapped to the generic representation, facilitating use of this data during research

Tests

Test (in a general sense) performed at a

certain moment in time

May be related to a

− Viral Isolate − Drug − Combination of a viral isolate and a drug

Adapts to “your (research) interest”

Tests vs Test Types

Test Type Concept Test Concrete Test Concrete Test Concrete Test Concrete

Tests vs Test Types

Viral Load Viral Load Generic Viral Load Abbot Viral Load “In house” Viral Load Roche

Test examples

Tests on a sample:

− CD4 count − Viral Load count

Tests on a sample with respect to a drug

− Therapeutic drug monitoring

Tests on a viral isolate

− HIV typing and subtyping

Tests on a viral isolate with respect to a drug

− Genotypic/phenotypic resistance test

Research datasets

Patient can be tagged by 1/more datasets User can have access to 1/more datasets Fine grained access to a Patient can be established User access towards a dataset can be

− anonymous-read-only (researchers)‏ − read-only − full read/write (clinicians)‏

Analysis Services

Built-in services

− REGA Subtype tool − Stanford ASI genotypic resistance algorithms

ANRS, HIVDB, REGA

− Resistance report generation

Using your own template documents

− Patient history visualisation

Integrate with external analysis services

− Using “web services”

Collaborative Edition

RegaDB Collaborative Edition Local RegaDB Local RegaDB Local RegaDB Local RegaDB

Collaborative Version

Easy and secure collaboration The sharing partner stays in control of his

data

− Access can be assigned in a very fine-grained

way

− Privileges can be removed when necessary

Same user environment as the local version

• f RegaDB

Research oriented features

Reproducibility

− A single version of a dataset is “immutable”

You can always rerun your analyses by going back to

the original patient group

Research tools

− Query tool to create queries

e.g. “Get one sequence per patient of patients failing

a specific treatment”

possibility to parametrize queries and store them in

RegaDB

dataset security on query level

− Export data to statistical / data mining packages

Synchronize auxiliary data

Auto-update services for

− Drug definitions − New (versions of) resistance algorithms − Test types / Tests − Attribute types / Attribute groups

RegaDB technologies

RegaDB is built on the foundations of

− Witty; a AJAX enabled Web Toolkit − Hibernate; a versatile DB layer

Currently we're extending RegaDB towards

• ther viruses

− HCV, HBV, influenza, ...

RegaDB usage scenarios

Data management at a local institute Collaboration based on the “Collaborative

Edition”

Data integration

− providing a local institute with clinical/research tools − providing data as a unified schema

ViroLab data integration

RegaDB is used to integrate ViroLab datasources

− locally installed at the institute − data conversion, extraction, and anonymization is done

within each hospital

ViroLab data integration

EUResist – ViroLab

HL7

− standard developed by IBM to communicate

clinical data

− a connector is under developement to convert

from RegaDB <-> HL7

Solution will ensure data origin of the

different partners

Thanks

Thanks for the invitation and your attention! Questions?