Latest Posters

Afitt is a software package for automated ligand conformation generation and placement within algorithmically identified unfilled electron density. Following real space refinement, the ligand solution is sent for subsequent refinement by Refmac or CNX, via coordinate and dictionary files. We have validated Afitt on forty publicly available data sets, chosen because it contains examples of highly strained ligand conformations.

We use local extrema in microarray time series data as the basis for discretisation. By comparing discretised gene expressions using similarity functions, we discover putative protein-to-protein interactions. We validate the results by use of public true positive and true negative databases for protein-to-protein interactions and demonstrate the high predictiveness of these local extrema as a time series feature.

Identification of low abundant proteins requires relative enrichment. We have tested systems for specific removal of high abundance proteins. Efficiency of albumin and IgG removal was assessed by ELISA and 2-DE. To estimate unspecific binding, six cytokines were spiked into serum. Recovery rates were determined with a cytometric bead assay.

Urine samples from healthy individuals and kidney cancer patients were used for metabolic profiling method development and for urine metabolome analysis.

A recently developed Metabolomics informatics solution from raw data processing to metabolite database searching, compatible with multiple instruments, has been successfully applied to the analysis of 1H NMR human serum spectra of diabetic and non-diabetic subjects. This study also highlights Overlap Density Heatmap, a novel technology that complements the use of chemometrics in biomarker identification.

Although it has been several years since metabolic networks became a commonly used analysis technique in bioinformatics, the question of how best to construct them from experimental data is still not satisfactorily resolved. Here we present a method for the construction of directed metabolic networks from microarray datasets using an enhanced version of the KEGG LIGAND database.

The novel quantitative technique of "Targeted Profiling" of 1H 1D NMR data is applied to a unique model of rheumatoid arthritis. A number of candidate biomarkers not previously identified using NMR techniques are identified related to inflammation and immune response. The pharmaceutical relevance of several of these metabolites is discussed.

At present, metabolite profiling is of growing importance in herbal medicine fields such as breeding, formulation, quality control and clinical trials. Herbal metabolic profiling allows direct detection of down-stream derivatives of metabolites, arising from herbal formulation process, using metabolite profiling ultra performance liquid chromatography time-of-flight mass spectrometry.

Metabolic profiling is concerned with the analysis of low molecular weight compounds present in complex samples such as plasma. Analytical techniques such as UPLC/MS are proving to be powerful tools for metabolic profiling on complex samples such as plasma. Here we describe a protocol of metabolic profiling analysis on plasma by UPLC/MS.

In the multivariate analysis of NMR-based metabolomics data, small variations in the resonance position of the individual peaks caused by experimental and instrument induced variations can adversely impact the PCA results. Techniques to address the NMR peak misalignment issue include the commonly used binning or bucketing, and the more advanced algorithms that aim at shifting the individual peaks to reach a better alignment across the spectra.