Current projects with bioinformatics-based inquiries:
1.
Protein sequence domains representing structural and functional
determinants of
enzymes’ active
sites are descriptors of physico-chemical composition associated with
their catalytic
activity. Therefore, it is hypothesized
that proteins sharing
such
physico-chemical descriptors may have reaction chemistry relatedness
even though they may fold into completely different 3-dimentional
structures. To test this hypothesis, we
integrate comparative
modeling of protein three-dimensional structures with phylogenetic
profiling to
find catalytic site similarity between proteins working in alternative
metabolic
processes, particularly in microbial enzymes supporting anaplerotic
pathways.
2.
Gene
ontologies and conceptual clustering methods are useful in exploring
gene-phenotype
associations. The approach considers that
knowledge of the biological role of genes and proteins in
one
organism can often be transferred to other organisms.
Therefore, disconnected pieces of information
that describe proteins’ molecular function, cellular location and
biological
process can be correlated using semantic-based bioinformatics tools to
find
plausible connections and develop and/or support hypotheses. In one application of this approach, we
reasoned a potential mechanistic relationship between folate transport
and
neural
tube closure defects by using gene ontologies. A
broader
implication of this line of work is to
address the challenge in integration of information in
gene/protein
databases and translating to practical
applications. The integrative and
translational aspect of information re-use can facilitate applications
critical
to biomedical research and healthcare industry, including personalized
medical
treatments, pharmaceutical drug repositioning, and medical informatics.
3.
Network theory is a powerful approach to study
organization principles of cellular components. Network
models can
help us understand the nature of fundamental connectivity in biological
pathways. It is
hypothesized that
stability, viability and agility of cells are dependent on the
compatibility
between interacting components. The interactions occur in a
flexible
network topology involving power law. Critical
disruptions and/or incompatibilities may cause large-scale collapse, or
loss of cellular
viability.
The project aims to examine this conjecture conceptually and establish
a model by using
matrix
analysis of a simple toy network. This model can
be used
to address the theoretical properties of a network consisting of a set
of
interactions between select proteins, which have ontological
relationships in
terms of biochemical function, process involvement and cellular
location. The approach is centered on the definition of a network,
which
considers any
collection of units potentially interacting as a system. Subtraction of
certain components or
addition of new
components as
well as perturbing changes in the existing components may alter the
connectivity and accessibility in biological networks, resulting in
variable
outcomes. Applications of this approach include understanding
evolutionary principles of protein interactions and correlating
physical
and
functional attributes of cellular components to biological processes,
such as metabolic pathways. Finding unique
characteristics of component connectivity and gaining insight into
emergence of
cellular responses that are caused by disturbed component
interactions, for instance, when microbial toxins exploit host
molecules,
can help us understand the basis of cellular
organization in physiological and pathophysiological conditions.
|