You are here:
Variability within Systemic In Vivo Toxicity Studies (ASCCT)
Citation:
Pham, Ly L., Woodrow Setzer, AND M. Martin. Variability within Systemic In Vivo Toxicity Studies (ASCCT). Presented at ASCCT, RTP, NC, September 29 - 30, 2016. https://doi.org/10.23645/epacomptox.5189353
Impact/Purpose:
Poster presentation at the ASCCT meeting in RTP, NC.
Description:
In vivo studies have long been considered the gold standard for toxicology screening. Often time models developed in silico and/or using in vitro data to estimate points of departures (POD) are compared to the in vivo data to benchmark and evaluate quality and goodness of fit. However, it is not certain what a good model is and how well is well before it turns into overfitting. Here we estimate the amount of variance that can be expected within systemic in vivo data. The present study was done using the US EPA’s Toxicity Reference Database (ToxRefDB). The database incorporates over 5,000 in vivo toxicity studies from the Office of Pesticide Programs (registrant-submitted studies), National Toxicology Program, pharmaceutical industries, and publically available literature covering over 1,000 chemicals. Using multilinear regression to calculate the residual sum or squares, we accounted for known variability in study conditions to quantify the unexplained variance of the log10(POD) to be about 0.35. The leave one out method was used to assess the amount of variance explained by each study condition and chemicals were found to be the biggest contributor. Stratifying the dataset by species and administration methods showed similar results, indicating stability of the unexplained variance. Considering and quantifying the unexplained variance will provide a benchmark and lower bounds on the mean-square-error for predictive toxicity model development.
URLs/Downloads:
DOI: Variability within Systemic In Vivo Toxicity Studies (ASCCT)
PHAM_POSTER_ASCCT_20162.PDF (PDF, NA pp, 331.829 KB, about PDF)
PHAM ASCCTABSTRACT2016_8-25 KMC.PDF (PDF, NA pp, 109.122 KB, about PDF)