ITF Health > Program > Speakers  > Wilfried Verachtert
linkedin twitter facebook email


Wilfried Verachtert    






Wilfried Verachtert graduated with a Master in Computer Science and started his professional career at the University of Brussels, where he conducted research on parallel programming languages. He then joined ITS, a spin-off where he started as a software developer of CASE tools and consultant for the introduction of software methodologies in larger companies. In 1998 he became a partner in MediaGeniX, a startup company working on innovative media asset systems for television stations in Europe. He joined imec in 2004 as Group Director MultiMedia, becoming Group Director for Digital Design Technology in 2005. Since 2010 he is the Director of the ExaScience Life Lab, a collaborative lab focussing on High Performance Computing in Life Sciences.



Presentation abstract


The wet lab cost of Whole human Genome Sequencing (WGS) has dropped faster than Moore's Law in the past 15 years and has now reached $1000, the threshold where massive clinical adoption becomes a reality. The resulting vast amount of genomic data enables new analytical algorithms. It will for example become possible to compare whole genomes of patients against a large population of other patients using advanced machine learning techniques. Screening large genomic databases – located at different centres –  for rare diseases then becomes feasible.

Such applications are complex, need access to non-centralized data and need large amounts of compute power. However, current bioinformatics ICT infrastructures, software, and algorithms needed to analyse and interpret genome information at scale have not followed the same trend as the wet lab costs. This raises the question whether the $1000 genome sequencing will not require $100,000 ICT analysis. We propose to use massively parallel software at scale to tackle this problem.

Besides the need for large amounts of compute power, preserving the locality of stakeholders’ sensitive information (including interoperability) poses another challenge for analytics encompassing multiple stakeholders. Simply pooling the stakeholders' resources has proven to be nearly impossible, with data silos, legal and privacy concerns and the impracticalities of moving large datasets around being the main hurdles. We propose the use of federated data analytics to tackle this problem.


Back to program

Back to speakers


Copyright imec 2017