analysis of genome-scale metabolic models has become a key
methodology to gain insights into functions, capabilities, and
properties of cellular metabolism. This systems biology tool has been
widely used in cancer research to predict potential vulnerabilities
in the metabolic network in the form of synthetic lethal. In brief,
synthetic lethals are sets of reactions/genes where only the
simultaneous removal of all reactions/genes compromises the viability
of the tumoral cell. However, the intertumoral heterogeneity between
patients )even with the same cancer type at the same stage)
represents an important challenge to be overcame in order to apply
these computational approaches in systems medicine approaches.
addition, since their inception, the size and complexity of
genome-scale metabolic reconstructions has significantly increase,
thus more computational resources are needed to analyze these
systems. This fact is enhanced by the exponential increase of
simulations required to unravel potential synthetic lethal
genes/reactions as a potential anti-tumoral targets.
the complexity and size of the metabolic networks together with the
large number of simulations needed for synthetic lethal analysis and
the inter-tumoral heterogeneity between patients with the same tumor
type and stage, make, in practice unfeasible an in-deep study of the
mechanisms underlying tumor progression and vulnerabilities via
model-driven methods as a translational tool in the scope of systems
it is necessary to develop a strategy to reduce the dimensionality of
the problem and find more effective ways to develop potential
multiple target treatments in complex and multi-factorial diseases
such as cancer.
project is aimed to develop a computational approach combining
multi-variate data analysis and model-driven methods that will allow
an in-deep discovery of multi-taget metabolic and gene regulatory
targets with potential anti-tumoral effects. More specifically it
will be achieved by applying three strategies:
metabolic model reduction and compaction
clustering in representative groups based on trascriptomic data
analysis in combination with algorithm for the rational reduction of
redundant simulations to detect pairs, triplets and quadruplets of
genes/reactiosn with anti-tumoral effects
two first approaches will drastically reduce the number of required
simulations while the third will permit to further analyze the
behavior of the system.
is expected that these computational approaches will enable a more
in-deep understanding of the complex molecular mechanisms underlying
tumor progression and the discovery of novel multi-target therapies
towards personalized medicine, that otherwise couldn’t be addressed
by current approaches.
project will be carried out in Prof. Lars Keld Nielsen lab (NNF-CFB
DTU, Denmark) and
will be under the direct supervision of Dr. Igor Marín.
successful appointee will
apply one or several of the strategies previously mentioned. Finally
the results will be analyzed and interpreted in order to describe the
mechanisms underlying tumor progression and the discovery of
potential novel multi-target therapies.
We are seeking for a highly motivated, independent, and well
organized person, who is passionate about computational biology.
Background on biostatistics and previous knowledge of some
programming language (R, Matlab, Python, ...) are desirable but not
students who are interested in join this project can contact to Igor
We are seeking for a highly motivated, independent, and well organized person, who is passionate about computational biology. Background in bioinformatics