Data fusion method

In general, different similarity measures yield different rankings, except when they are monotonic. Improved results are obtained by using data fusion methods to combine the rankings resulting from different coefficients. 

Whittle, M., Gillet, V.J., Willett, P., and Loesel, J. (2006) Analysis of data fusion methods in virtual screening theoretical model. Journal of Chemical Information and Modeling, 46 (6), 2193-2205. 

Chemometrican Data management and data fusion Process data analysis Multivariate data analysis Analyzer calibration model development Method equivalence Process models development (e.g., MSPC) Experimental design (e.g., DOE)... 

It is recommended to use data fusion techniques such as the Fuzzy approach or other methods like the Neuro-Fuzzy on surface data to locate the most promising sites for drilling. 

Hert J, WiUett P, Wilton DJ, Addin P, Azzaoui K, Jacoby E, Schuffenhauer A. (2005) New Methods for Ligand-Based Virtual Screening Use of Data-Fusion and Machine-Learning Techniques to Enhance the Effectiveness of Similarity Searching. /. Chem. Inf. Model, (in the press). 

Although detectable concentrations for several elements could be found after fusion, it is felt that the volatility of mercury and possibly lead and tin would make their determination by lithium tetraborate fusion questionable. Table I shows the elements selected for analysis and the accuracy and precision data for the standards used to check the fusion method. Each standard in Table I was of known composition and siliceous in nature. The standards were separately prepared 10 times so that a statistical evaluation of the results could be made. The standards used were USGS Standards G-2, W-l, BCR-1, commercially prepared silica-alumina based standards, and unfused synthetic standards prepared by the Coal Research Bureau (9, 10, 11, 12). The synthetic standards were used because no commercially prepared standard having... 

The chemical compositions of the ancient Egyptian Blue samples (reported in the following section) were determined by atomic absorption spectrophotometry using the hydrofluoric acid digestion method together with the lithium metaborate fusion method for the silica determination (9). Some 20-30 mg of powder drilled from the objects was used for these analyses. Additionally, the arsenic concentrations were determined by x-ray fluorescence spectrometry. The precision of the analytical data was 1-2% for the major elements (>10% concentration) and deteriorated to 5-20% for the trace elements (<0.1% concentrations). However, due to the inhomogeneity of the material, variations in elemental concentrations (i.e., major, minor, and trace) of 10-15% can be expected within a single object. 

We focus on the combination of transcriptomics and metabolomics and more specifically on microarray data, which is currently the most used method for gene expression profiling and is used on a routine basis. In the first paragraphs, we briefly revise the extraction of mRNA or metabolites, their measurement, quality control of data, and analysis methods. Afterward two different types of data fusion and recent tools and publications are reviewed, followed by visualization methods for obtained data. Lastly, the metabolite annotation Web server MassTRIX is presented. This Web server allows combined analysis of transcriptomic and metabolomic data in the context of metabolic pathways. We compared the metabolomics part against similar tools and give a short outlook on the next version of MassTRIX, MassTRIX 4. 

These are just three examples of low-level data integration showing the capabilities of this approach. In the above-mentioned articles, previously known metabolite-transcripts were found along novel ones. In most cases, metabolites are known, but such analysis can be even further developed for the biological interpretation of unknown molecules. The major advantage of low-level data fusion is that a priori no knowledge about the studied system is needed, although for method validation known associations are needed. 

The optical microscope is a valuable tool in the laboratory and has numerous applications in most industries. Depending on the type of data that is required to solve a particular problem, optical microscopy can provide information on particle size, particle morphology, color, appearance, birefringence, etc. There are many accessories and techniques for optical microscopy that may be employed for the characterization of the physical properties of materials and the identification of unknowns, etc. Utilization of a hot-stage accessory on the microscope for the characterization of materials, including pharmaceutical solids (drug substances, excipients, formulations, etc.), can be extremely valuable. As with any instrument, there are many experimental conditions and techniques for the hot-stage microscope that may be used to collect different types of data. Often, various microscope objectives, optical filters, ramp rates, immersion media, sample preparation techniques, microchemical tests, fusion methods, etc., can be utilized. 

The combination of virtual screening results with several reference ligands and/or several search methods is called data fusion. Whittle et al. [49, 50] refer to the combination of similarity searches for one reference ligand with different descriptors as similarity fusion, whereas the combination of the results for a set of reference ligands with one method is called group fusion. Many reports on different methods and results of data fusion approaches can be found in the literature [41, 46, 51-53]. 

Most researchers agree that data fusion is insensitive to the presence of a poorly performing method. This leads to a more consistent level of search performance compared to individual methods, because the performance of individual methods varies from target to target in an unpredictable manner. 

Schuffenhauer et al. [21] reported a comparative study ofbioisosteric replacements with UNITY 2D fingerprints and FBSS (field-based similarity search) using the Bioster database [22] as a source ofbioisosteric pairs. The authors report that both these 2D and 3D methods provide complementary results that were demonstrated to work synergistically when combined using data fusion. The UNITY fingerprint was reported to be very sensitive to heteroatom replacement. This sensitivity can be overcome somewhat by abstracting the atoms present in a structure into pharmacophoric features. 

Hert J, Willett P, Wilton DJ, et al. New methods for ligand-based virtual screening Use of data fusion and machine learning to enhance the effectiveness of similarity searching. J Chem Info Model 2006 46 462-470. 

Section 15.4 provides a discussion of similarity measures, which depend on three factors (1) the representation used to encode the desired molecular and chemical information, (2) whether and how much information is weighted, and (3) the similarity function (sometimes called the similarity coefficient) that maps the set of ordered pairs of representations onto the unit interval of the real line. Each of these factors is discussed in separate subsections. Section 15.5 presents a discussion of a number of questions that address significant issues associated with MSA Does asymmetric similarity have a role to play Do two-dimensional (2D) similarity methods perform better than three-dimensional (3D) methods Do data fusion and consensus similarity methods exhibit improved results Are different similarity measures statistically independent How do we compare similarity methods Can similarity measures be validated S ection 15.6 provides a discussion of activity landscapes... 

All of this suggests that when aggregating similarity scores by data fusion or related methods, the scores produced by all of the methods should be considered unless the methods generating the scores are too closely related with respect to the representations and similarity functions being used. 

Both oxygen and carbon were determined by the inert gas fusion method which is quite useful in measuring small amounts of these elements. Tie data show no difference in the average oxygen content of the riffled vs. the unriffled samples. Tlie carbon content is typical of the silicon nitride powders prepared by the vendor s method. 

A new approach to similarity data fusion, based on belief theory, yields some interesting methods for performing ligand-based screening. The best... 

Ren and Gao [27] developed a technique which combines data fusion (DF) and multiscale wavelet transformations with generalized regression neural network (GRNN) and applied it for analyzing overlapping spectra. The main role of the hybrid method, DF-GRNN, is to enhance... 

---