One of the strategies of the Francis Crick Institute is to
collaborate creatively, taking a multi- and inter-disciplinary
approach. Some recent work that has been facilitated by
the Crick ICT team and the astronomy and physics groups at UCL and
Durham Universities shows how this is can work in
action.
Alessandro Costa is working on the architecture and dynamics of
macromolecular machines at the Cancer Research UK London Research
Institute and the Francis Crick Institute. His group has been
piloting running large scale compute jobs on the DiRAC (Distributed
Research utilising Advanced Computing) supercomputing
infrastructure. The mission of the DiRAC resource is :
"DiRAC is the integrated supercomputing facility for theoretical
modelling and HPC-based research in particle physics, astronomy and
cosmology, areas in which the UK is world-leading. HPC-based
modeling remains an essential tool for the exploitation of
observational and experimental facilities in astronomy and particle
physics". However, a proportion of the capacity (10%) is
reserved for non-physics problems in order to promote
inter-disciplinary use of this capability.
In order to prove the feasibility of running his calculations on
the infrastructure Alessandro's group has already used 28,000 CPU
hours and expect this to reach 2.5 million CPU hours per year when
they reach production scale. This would constitute 10% of the
capacity at Durham, which means that LRI and the Crick are
benefitting from all of the capacity there which is available for
non-physics research.
This is a new departure for much of life sciences research in
developing the use of the kind of supercomputing resource which had
previously been the preserve of space science, high energy physics
and earth sciences. Biologists are getting an advantage from the
experience and expertise that are located in resources such as
DiRAC to rapidly gain access to the kinds of resource that they now
require for data synthesis and modelling.
The Durham DiRAC site noted the high level of technical
competence of the LRI/Crick group and the ease with which their
code has been able to run on the infrastructure. Similarly the
biomedical scientists have been happy with the supportive and
helpful relationship with Durham and DiRAC.
Alessandro describes his work as addressing cancer by "A
hallmark of cancer cells is genomic instability, arising from
errors in the mechanisms that maintain gene copy number and
chromosome ploidy. Our research aims to understand how
macromolecular machines involved in DNA replication function to
preserve chromosome integrity. To address these issues, we employ a
combination of single particle cryo-electron microscopy, molecular
modelling and biochemistry, to generate mechanistic models that
explain the basis of key nucleic acid transactions; for example, we
are interested in understanding how replicative helicases
collaborate with polymerases to couple DNA unwinding with DNA
synthesis. By describing the architecture and dynamics of the DNA
replication machinery, we seek to establish a molecular framework
that explains how higher eukaryotes respond to DNA damage and how
cell proliferation is regulated to avoid tumorigenesis."