Related Experiment Video
Updated: Feb 8, 2026

Physical, Chemical and Biological Characterization of Six Biochars Produced for the Remediation of Contaminated Sites
Published on: November 28, 2014
Characterizing the effects of missing data and evaluating imputation methods for chemical prioritization applications
Kimberly T To1, Rebecca C Fry2, David M Reif1,3,4
11Bioinformatics Research Center, North Carolina State University, 1 Lampe Dr, Raleigh, 27695 NC USA.
Missing data in chemical assessments significantly impacts prioritization scores. Simulation studies reveal that imputation methods alter chemical rankings, especially with limited data, highlighting the need for careful strategy selection in toxicological profiling.
Area of Science:
- Environmental Science
- Toxicology
- Computational Chemistry
Background:
- The Toxicological Priority Index (ToxPi) is a chemical prioritization and profiling method integrating diverse data sources.
- Missing data in chemical assays (e.g., in vitro, in vivo) is common, where not all chemicals are tested in every assay.
- The Agency for Toxic Substances and Disease Registry's (ATSDR) Substance Priority List (SPL) identifies chemicals for environmental research and remediation.
Purpose of the Study:
- To investigate the effects of missing data on chemical prioritization using simulation studies.
- To evaluate various data imputation methods for addressing missing data in high-throughput screening (HTS) datasets.
- To recommend solutions for handling missing data in toxicological profiling.
Main Methods:
- Designed simulation studies using HTS data from ToxCast and Tox21 programs.
- Explored scenarios with varying percentages of missing assay data (0-80%) per chemical.
- Tested multiple imputation methods including k-Nearest-Neighbor (kNN), Singular Value Decomposition (SVD), Max, Mean, Min, Binomial, and Local Least Squares.
Main Results:
- Most imputation methods significantly altered ToxPi scores, except in datasets with few assays.
- Chemical ranks were found to be more sensitive to score changes with minimum value imputation, SVD imputation, and kNN imputation.
- The impact of imputation on scores and ranks was most pronounced in scenarios with fewer assays and higher proportions of missing data.
Conclusions:
- The choice of imputation strategy significantly influences chemical scores and their associated ranks in toxicological profiling.
- Understanding the effects of missing data and the performance of imputation methods is crucial for robust decision-making.
- Characterizing these effects enhances confidence in assessing the health and ecological impacts of environmental chemicals.
Related Concept Videos
Effects of Chemicals: Overview
Methods of Sterilization II: Chemical Methods
Using chemical sterilization rather than heat to clean out equipment is recommended. It eradicates and removes all bacteria,...
Inductive Effects on Chemical Shift: Overview
Ionic Strength: Effects on Chemical Equilibria
In this solution, the primary...
Chemical Shift: Internal References and Solvent Effects
The internal reference compound generally used in NMR spectroscopy is tetramethylsilane (TMS). TMS is preferred because it is chemically inert, soluble in NMR solvents, and easily removable. Also, the highly shielded methyl protons in TMS yield an intense...
π Electron Effects on Chemical Shift: Overview

