Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Discriminant function analyses of liver-specific carcinogens.

Richard D Beger1, John F Young, Hong Fang

  • 1Division of Chemistry, Food & Drug Administration, National Center for Toxicological Research, Jefferson, AR 72079, USA. rbeger@nctr.fda.gov

Journal of Chemical Information and Computer Sciences
|May 25, 2004
PubMed
Summary

Predicting chemical carcinogenicity aids FDA reviews. A new liver cancer database (NCTRlcdb) and partial least squares discriminant function (PLS-DF) modeling show promise, though sensitivity for liver cancer prediction needs improvement.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Randomized, Placebo-Controlled, Multicenter Clinical Study on the Efficacy and Safety of Lidocaine Patches in Chinese Patients with Postherpetic Neuralgia.

Dermatology and therapy·2023
Same author

Rare tumors: a blue ocean of investigation.

Frontiers of medicine·2023
Same author

Treatment of primary cutaneous lichenoid amyloidosis with abrocitinib: A pilot study in two cases.

International journal of dermatology·2023
Same author

Adverse events of PD-(L)1 inhibitors plus anti-VEGF(R) agents compared with PD-(L)1 inhibitors alone for cancer patients: a systematic review and meta-analysis.

Frontiers in pharmacology·2023
Same author

CircHIPK2 facilitates phenotypic switching of vascular smooth muscle cells in hypertension.

Journal of human hypertension·2023
Same author

Clinical Features and Skin Microbiome of Tinea Scrotum: An Observational Study of 113 Cases in China.

Mycopathologia·2023

Area of Science:

  • Toxicology
  • Computational Chemistry
  • Drug Safety

Background:

  • Predicting organ-specific carcinogenicity is crucial for evaluating new chemical applications by regulatory bodies like the FDA.
  • Existing methods for carcinogenicity assessment can be time-consuming and resource-intensive.
  • A comprehensive database of chemical compounds with associated carcinogenicity data is needed for developing predictive models.

Purpose of the Study:

  • To develop and validate a predictive model for liver-specific carcinogenicity using a curated database of chemical compounds.
  • To assess the performance of partial least squares principal component discriminant function (PLS-DF) modeling in predicting liver cancer.
  • To provide tools for FDA reviewers to aid in the evaluation of new chemical applications.

Main Methods:

Related Experiment Videos

  • Construction of the NCTR liver cancer database (NCTRlcdb) comprising 999 compounds with Cerius2, Molconn-Z, and (13)C NMR descriptors.
  • Classification of each compound in the NCTRlcdb as either liver cancer or non-liver cancer.
  • Application of partial least squares principal component discriminant function (PLS-DF) modeling to evaluate liver-specific carcinogenicity.

Main Results:

  • PLS-DF models using a priori probabilities of 0.29 (liver cancer) and 0.71 (noncancer) achieved 70.6% overall predictability, with 18.8% sensitivity for liver cancer and 90.8% specificity for noncancer.
  • PLS-DF models using equal a priori probabilities (0.50 for both classes) resulted in 61.0% overall predictability, with 50.5% sensitivity for liver cancer and 65.3% specificity for noncancer.
  • The results indicate that model performance is sensitive to the choice of a priori classification probabilities.

Conclusions:

  • The NCTRlcdb and PLS-DF modeling approach offer a potential method for predicting organ-specific carcinogenicity.
  • The developed models demonstrate moderate overall predictability but highlight challenges in achieving high sensitivity for liver cancer prediction.
  • Further refinement of modeling strategies and descriptor sets may be necessary to improve the accuracy and reliability of carcinogenicity predictions for regulatory applications.