EuResist Network Present and future Francesca I ncardona I I I - - PowerPoint PPT Presentation
EuResist Network Present and future Francesca I ncardona I I I - - PowerPoint PPT Presentation
EuResist Network Present and future Francesca I ncardona I I I EuResist I nternational meeting Bonn 4 May 2012 The EuResist EU project: 2006-2008 Topics and results Integration of viral genomics with clinical data to predict response to anti
III EuResist International meeting Bonn 4 may 2012
The EuResist EU project: 2006-2008 Topics and results
Integration of viral genomics with clinical data to predict response to anti HIV treatment The EuResist Integrated DataBase
(EIDB): the largest existing database of Genotype-response information: more than 33.000 patients
The EuResist Prediction
engine: accuracy of 76% Favourable comparison with state of the art prediction systems and… with humans: the EVE study
Type of Data EuResist Patients
33.832
Therapies
98.421
CD4 Isolates
455.669
Viral Load Isolates
372.202
Sequences
30.938
III EuResist International meeting Bonn 4 may 2012
The EuResist Network GEI E
A European no-profit grouping Founded by Informa, Karolinska Institute, Max Planck Institute, University of Cologne, University of Siena Manages project results and brings them forward Offers partnership to new parties of proved or potential scientific value All partners become members of the EuResist Network Scientific Board The EuResist Network Scientific Board evaluates the scientific adequacy of research proposals that require the access to the data stored in the EuResist Integrated Data Base (EIDB)
III EuResist International meeting Bonn 4 may 2012
Partnership with EuResist Network
Reasons for partnership include either expanding the EIDB with valuable data or contributing to data analysis in scope with the EuResist Network aims or both. Partner’s activities and data are protected by the data and authorship policies Data provided by a partner always remain the property of the partner. Data providers are included in the list of authors. The priority list of data providers to be included as authors reflects the proportion of contributed cases in the study and credits are accumulated by data providers not included (as in ARCA).
III EuResist International meeting Bonn 4 may 2012
Partners
InformaPRO Italy Karolinska inst. Sweden Max Planck Germany UniKoeln Germany UniSiena Italy Portugal Irsicaixa Spain Luxembourg KuLeuven Belgium World Friends Kenya MUHAS Tanzania NRL Rwanda Hungary
III EuResist International meeting Bonn 4 may 2012
Update of the EI DB structure
August 2011 The new schema allows to store: new HIV genes new species such as HCV and HBV New tables allow to store other information such as: infections, diagnoses, type of drugs
III EuResist International meeting Bonn 4 may 2012
The EuResist I ntegrated DataBase EI DB
III EuResist International meeting Bonn 4 may 2012
Present EI DB coverage
KENYA PORTUGAL
ITALY
III EuResist International meeting Bonn 4 may 2012
EI DB
Patients per Centrum Tot 60856
10416 8742 8374 1361 1119 1003 865 657 200 28119 5000 10000 15000 20000 25000 30000 Arca Arevir Karolinska Lisbon BI Leuven Irsicaixa Rwanda Luxemburg Tibotec
III EuResist International meeting Bonn 4 may 2012
EI DB
Standard Datum Tot 9857
4473 1484 1419 1166 436 361 228 189 101 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Arca Arevir Karolinska BI Luxembourg Leuven Lisbon Irsicaixa Tibotec Rwanda
III EuResist International meeting Bonn 4 may 2012
EI DB
Gender Distribution (60.829 patients)
Male 63% Female 25% Unknown 12%
Undifferentiated 0%
III EuResist International meeting Bonn 4 may 2012
EI DB
Age Distribution (55.798 patients)
36-45 31% 45< 53% 0-13 1% 14-25 2% 26-35 13%
III EuResist International meeting Bonn 4 may 2012
EI DB
Ethnic Groups Distribution (58.961 patients)
Asian 1% Caucasian 20% Unknown 72% Afro 6% Other 0% Hispanic 1%
III EuResist International meeting Bonn 4 may 2012
EI DB
Risk Factor Distribution (58.600 patients)
Ivda 12% unknown 35% blood products 1% vertical transmission 1% heterosexual 19%
homosexua bisexual 19%
- ther
13%
III EuResist International meeting Bonn 4 may 2012
EuResist Engine usage May 2011 – Apr 2012
III EuResist International meeting Bonn 4 may 2012
EuResist Network website usage May 2011 – Apr 2012
III EuResist International meeting Bonn 4 may 2012
I ntercontinental activities
EDCTP ENNEA project Kenya, Tanzania, Ethiopia PLANTA project Central America SISGENO-EURESIST comparison Brasil Other affiliated projects DOTARV: Vietnam HIVIND: India Eastern promises: Russia (n= 3)
III EuResist International meeting Bonn 4 may 2012
Other ongoing activities
CHAIN: Collaborative HIV and anti-HIV drug resistance network - EU Research project EUCOHIV project: Collection of Maraviroc treatment cases – sponsored by ViiV Collaboration with Stanford for the realisation of a freely accessible genotype-response repository CETRA project: Collection of Etravirine treatment cases
- sponsored by Tibotec
EuResist Engine: refinement of prediction error estimate Retraining of EuResist Engine
III EuResist International meeting Bonn 4 may 2012
Can we rise prediction accuracy?
EuResist Engine trained on 3000 TCE: AUC = 0,76 EuResist Engine trained on 5000 TCE: AUC = 0,80 RDI engine trained on 5.700 TCE: AUC= 0,82
Revell et al AIDS 2011
It seems there is a plateau in accuracy prediction What can we do? EuResist Engine will be trained on new EIDB Take into account human genomics
III EuResist International meeting Bonn 4 may 2012
Human genomics: the next step
The pattern of Human Leukocyte Antigen (HLA) alleles or supertypes is known to impact HIV disease progression Recent genome-wide association studies (GWAS) done in HIV-positive patients have identified a few single nucleotide polymorphisms (SNPs) partly explaining individual differences in the level of spontaneous control of HIV replication The inter-host network of molecular interactions variability may allow for prediction of anti-HIV therapy outcome and even for personalized treatment
Tang 2003, Singh 2008, Tang 2010, Arab-Alameddine 2011, An 2010, Rotger 2010, van Manen 2011, Fellay 2009, Fellay 2010, van Manen 2011
III EuResist International meeting Bonn 4 may 2012
Multilevel virus-host interaction modeling
Integrating the host genomics Integrating models at different levels (from molecular to clinical to population level) Integrating machine learning prediction models with biomechanistic description models Collaboration started with DTU, Stanford University, Florida University
III EuResist International meeting Bonn 4 may 2012
Objectives for the future
To be able to provide treatment support in all countries with relevant HIV infected population To upgrade the prediction system with possibility to predict without genotype (Prosperi et al. PLoSone
2010)
To upgrade the prediction system taking into account patient’s immune system To make a HCV EuResist
III EuResist International meeting Bonn 4 may 2012
I n the very near future
To strenghten the intra-Network communication for: awareness, collaboration, scientific ideas circulation:
3 tele-meetings per year: March, June November
To better exploit the EIDB:
A person for querying it systematically and propose new studies You are all invited to propose studies on the EIDB!
III EuResist International meeting Bonn 4 may 2012
Thank you
24