Research Activities: Bioinformatics

The Bioinformatics group of our lab is mainly focused on the development of algorithms that predict the structure, function and interaction of proteins. These algorithms frequently have genome-wide applications. More specifically:

Prediction of protein structure, function and interactions
Our research is mainly focused on developing algorithms that are capable of predicting the structure, function and interactions of proteins. We are, in particular, focusing our efforts on the topology prediction of transmembrane proteins, either alpha-helical ones (PRED-TMR,CoPreThi, orienTM, waveTM) or beta-barrels (PRED-TMBB, ConBBPred). Furthermore, we developed algorithms for structural and/or functional classification of proteins readily available for genome-wide applications. Such applications are available for structural classification of proteins (PRED-TMR2, PRED-CLASS, MCMBB) and for classification of the G-protein Coupled Receptors into families (PRED-GPCR), or according to the G-protein coupling specificity (PRED-COUPLE, PRED-COUPLE2). Towards these goals, we use sophisticated mathematical and computational methods ranging from various methods of statistical analysis, to Neural Networks, Support Vector Machines, Dynamic Programming and Hidden Markov Models.

Development and annotation of biological databases
The knowledge expertise of our team, in combination with the sophisticated tools that we develop, enables us to build curated, expert, knowledge-based, relational biological databases. These databases may be used either by bioinformaticians or by molecular biologists. We have been focused, until now, mainly on biologically interesting classes of proteins that, either are not fully annotated in publicly available databases, such as the Outer Membrane Proteins (OMPs) of Gram-negative bacteria and the insect cuticle proteins (cuticleDB), or to classes, with important functional roles that involve protein-protein or protein-nucleic acid interactions (gpDB, DNA-B).

Sequence alignment, Periodicities, and low complexity regions discovery
Our research team focuses also on discovering hidden periodicities in protein and DNA sequences, using either the Fourier Transform (FT) or the Wavelet Transform (waveTM). We are also studying the low complexity regions in protein and nucleic acid sequences, and we developed algorithms for filtering such sequences (CAST). Using pairwise alignments, we implement algorithms for clustering large datasets and create non-redundant sets (NON-RED). Recently, research has been focused on improving the pairwise and multiple alignment algorithms quality, by either incorporating structural or evolutionary information (unpublished). Furthermore, tools for the represantation of membrane protein structure were also developed (TMRPres2D), as well as tools capable of visualizing gene-product functional and structural features in genomic datasets (GeneViTo).

Research Activities: Biophysics
The scientific research interests of our laboratory are also oriented towards the field of structural biology and molecular biophysics. More specifically, we are focused on:

Structural studies of peptide-analogues parts of silkmoth chorion proteins, as novel self-assembled polymers with amyloid properties.
Chorion is the major component of silkmoth eggshell. More than 95% of its dry mass consists of proteins that have remarkable mechanical and chemical properties protecting the oocyte and the developing embryo from a wide range of environmental hazards. Synthesized peptide-analogues of parts of chorion proteins fold and self-assemble forming amyloid-like fibrils, under a wide variety of diverse conditions. This raises the question whether chorion is a natural, protective amyloid. The folding and self-assembly mechanisms of these peptides are being extensively studied utilizing electron microscopy (negative staining and shadowing), X-ray diffraction, FT-Raman, ATR FT-IR and CD spectroscopy and computer modelling. Principles that govern the self-assembly of proteins into amyloid-like structures are being unravelled and may be important in a variety of pathological cases in amyloidoses like Alzheimer’s, transmissible spongiform encephalopathies (mad-cow disease, Creutzfeld-Jacobs, prion diseases etc.), type II diabetes etc.

Studies of protein-chitin interactions for the formation of arthropod cuticle.
Cuticle is a complex, bipartite structure, composed mainly of proteins and chitin, which provides protective, locomotive and structural functions important for arthropod survival. Chitin-protein interactions are studied utilizing X-ray diffraction, FT-Raman, ATR FT-IR and CD spectroscopy of natural and specially treated samples. Cuticle protein sequence alignment, secondary structure prediction, computer modelling and docking studies are also used. Recently, a relational cuticle protein database (cuticleDB), was also constructed in our lab and is freely provided through the internet. Also, recently, a model for the structural proteins both of hard and soft cuticles has been proposed and guides our efforts towards unravelling the molecular principles that dictate cuticle overall molecular architecture.

Structural and self-assembly studies of fibrous proteins, which form structures of physiological importance like silkmoth and fish chorion.
Silkmoth and fish chorion is a helicoidal composite (biological analogue of a cholesteric liquid crystal) of protein fibres embedded in a protein matrix. The principles governing the self-assembly of chorion protein molecules into helicoidal proteinaceous extracellular structures are being studied. X-ray diffraction, FT-Raman, ATR FT-IR spectroscopy, electron microscopy and computer modelling are the main techniques used to achieve this goal.

Structural studies of protein molecules that play significant roles in cell functions, through determination of protein structures (and as complexes with small organic compounds) utilizing X-ray crystallographic methods.
Research is carried out on the enzyme Dihydrofolate Reductase (DHFR), which participates in a biological pathway that leads to the formation of thymine which is essential in DNA biosynthesis and various aminoacids. Blockade of the action of this enzyme leads to cell death. In the context of anticancer and antimicrobial research, DHFR has extensively been studied. Our goal is to study the interaction of DHFR with small organic compounds, inhibitors of the enzyme in order to design compounds appropriate for clinical treatment.
Furthermore, efforts are centered on the study of Concanavalin A (Con A), which is a representative member of the lectin class of plant proteins. It generally exhibits specificity for saccharides containing a-D-mannose or a-D-glucose residues, but it may also bind oligosaccharide sequences lacking these units. Con A has specific biological activities which depend on its binding to cell surface receptors. It agglutinates cells transformed by oncogenic viruses, inhibits growth of malignant cells in animals, and exhibits mitogenic activity. Although the exact biological role of Con A still remains unknown, its specific saccharide-binding properties make it an ideal object for the study of protein–saccharide interactions.