Methods for Data Analysis

In science in general, and in chemistry and in pharmaceutical research in particular, huge amounts of data are produced. The intrinsic information in these data is often difficult to grasp. In many cases, not only are the data themselves interesting, but the intrinsic relationships among the data are of particular interest. 

Chemical data contain information about various characteristics of chemical compounds and a wide spectrum of methods are applied to extract the relevant information from the data sets. Data analysis, however, not only deals with the extraction of primary information from data but also with the generation of secondary 

Chemoinformatics A Textbook. Edited by Johann Gasieiger and Thomas Engel Copyright 2003 Wiley-VCH Verlag GmbH Co. KGaA. 

---

Figure 3-17. Sohematios of the differential method for data analysis.

These problems were addressed by Tidwell and Mortimer117 118 who advocated numerical analysis by non-linear least squares and Kelen and Tiidos110 1"0 who proposed an improved graphical method for data analysis. The Kelen-Tiidos equation is as follows (eq. 43) ... 

Sec. 820.50 Statistical techniques - Use appropriate statistical methods for data analysis and... 

The overall objective of the system is to map from three types of numeric input process data into, generally, one to three root causes out of the possible 300. The data available include numeric information from sensors, product-specific numeric information such as molecular weight and area under peak from gel permeation chromatography (GPC) analysis of the product, and additional information from the GPC in the form of variances in expected shapes of traces. The plant also uses univariate statistical methods for data analysis of numeric product information. 

Chambers, J.M., Cleveland, W.S., Kliner, B. and Tukey, PA. (1983). Graphical Methods for Data Analysis. Duxbury Press, Boston. 

Carlin BP, Louis TA. 2000. Bayes and empirical Bayes methods for data analysis. 2nd ed. Boca Raton (FL) Chapman and Hall/CRC. 

The experimental setup and the testing procedures specifically needed for the dynamic kinetic investigation will be discussed in the following sections, as well as the relevant methods for data analysis and for parameter estimation. 

Biomedical research at the beginning of the century usually involved relatively few measurements on many subjects. While measurement was expensive, the cost for experimental subjects was often low. The methods for data analysis developed during the same period - r-tests, analysis of variance, multiple regression - were optimized for this situation with many subjects and few measurements. 

Today the situation is the reverse. Multiple measurements are often made on each subject. By taking as many measurements as possible, it is hoped that the complete picture of each subject will be less error-prone than each individual measurement considered separately. As an illustration, the development of chromatographic separation methods and detection techniques like spectrophotometry has made it possible to identify many substances in one blood sample from a patient. At the same time, the cost of each experiment has increased. Consequently, new methods for data analysis are needed, methods that can utilize the information in the data also when the subjects are few and the measurements are many. 

Now that data have been stored, we come to analysis, which is the real point of everything we have done so far. Methods for data analysis vary. They vary with the inclinations of the software programmer they also vary with the budget of the cytometer facility. They may be strictly commercial, or they may be homemade. These days, commercial manufacturers of cytometers compete with each other on the basis of their software systems as much as on the basis of their cytometer technology. In addition, independent entrepreneurs, with increasing frequency, have begun to program for analysis of... 

To design an experiment means to choose the optimal experiment design to be used simultaneously for varying all the analyzed factors. By designing an experiment one gets more precise data and more complete information on a studied phenomenon with a minimal number of experiments and the lowest possible material costs. The development of statistical methods for data analysis, combined with development of computers, has revolutionized the research and development work in all domains of human activities. 

The EPA developed a document titled Guidance for Data Quality Assessment, Practical Methods for Data Analysis, EPA QA/G-9 (EPA, 1997a) as a tool for project teams for assessing the type, quality, and quantity of data collected for projects under the EPA oversight. This document summarizes a variety of statistical analysis techniques and is used primarily by statisticians. DQA, however, is not just a statistical evaluation of the collected data. It is a broad assessment of the data in the context of the project DQOs and the intended use of the data, which requires a... 

US Environmental Protection Agency, Guidance for Data Quality Assessment, Practical Method for Data Analysis, EPA QA/G-9, [US Environmental Protection Agency, 1997a]. 

Ruzic, I. and Pecar,J. (1997) A new method for data analysis in studies of trace metal com-plexation. Croat. Chem. Acta, 70, 229-245. 

The Lateral Ignition and Flame spread Test (LIFT) apparatus was developed primarily for lateral flame spread measurements. The apparatus, test procedures, and methods for data analysis are described in ASTM E 1321. A sample of 155 x 800 mm is exposed to the radiant heat of a gas-tired panel. The panel measures 280 x 483 mm. The heat flux is not uniform over the specimen, but varies along the long axis as a function of distance from the hot end as shown in Figure 14.6. The flux distribution is an invariant of distance when normalized to the heat flux at the 50 mm position. When methane or natural gas is burnt, the upper limit of the radiant heat flux is 60-65 kW/m2. The lower limit is approximately 10kW/m2 since the porous ceramic tile surface of the panel is only partly covered with flame at lower heat fluxes. 

USEPA (2000) Guidance for data quality assessment. Practical methods for data analysis. Washington, DC, United States Environmental Protection Agency, Office of Research and Development, July (EPA/600/R-96/084 http //www.epa.gov/quality/qs-docs/g9-final.pdf). 

In part, the widespread and routine use of single-crystal X-ray diffraction techniques in the modern day is a result not only of advances in instrumentation but also of the development of tremendously powerful methods for data analysis, such that, in the vast majority of cases, the crystal structure can be determined almost routinely even from experimental data of only modest quality. Thus, provided a single crystal of sufficient size and quality is available for the material of interest, successful structure determination by analysis of single-crystal X-ray diffraction data is nowadays a virtual formality. 

---