Data quality in predictive toxicology: reproducibility of rodent carcinogenicity experiments.

We compared 121 replicate rodent carcinogenicity assays from the two parts (National Cancer Institute/National Toxicology Program and literature) of the Carcinogenic Potency Database (CPDB) to estimate the reliability of these experiments. We estimated a concordance of 57% between the overall rodent carcinogenicity classifications from both sources. This value did not improve substantially when additional biologic information (species, sex, strain, target organs) was considered. These results indicate that rodent carcinogenicity assays are much less reproducible than previously expected, an effect that should be considered in the development of structure-activity relationship models and the risk assessment process.

Data quality in predictive toxicology: identification of chemical structures and calculation of chemical properties.

Every technique for toxicity prediction and for the detection of structure-activity relationships relies on the accurate estimation and representation of chemical and toxicologic properties. In this paper we discuss the potential sources of errors associated with the identification of compounds, the representation of their structures, and the calculation of chemical descriptors. It is based on a case study where machine learning techniques were applied to data from noncongeneric compounds and a complex toxicologic end point (carcinogenicity). We propose methods applicable to the routine quality control of large chemical datasets, but our main intention is to raise awareness about this topic and to open a discussion about quality assurance in predictive toxicology. The accuracy and reproducibility of toxicity data will be reported in another paper.

Evaluating antiarrhythmic strategies: a knowledge-based system for exploring clinical data.

Medical therapy for cardiac arrhythmias is still to a large extent based on empirical methods. Assessing and evaluating different therapeutical strategies constitutes the starting point for inducing decision methods to select the appropriate regimen for an individual patient. We designed a computer-based system that establishes a set of heuristic rules linking attributes in a data base of patients with rhythm disturbances. A feasibility analysis conducted on a small set of 23 patients indicated that constraints on the number of attributes and their clinical relevancy together with a representation scheme for temporal changes have to be incorporated to provide for a useful and efficient algorithm.