Skip to main content
An official website of the United States government
Email

History of the NCI-60 screen and COMPARE algorithm

The COMPARE algorithm is a powerful tool developed by NCI to help researchers explore and interpret patterns of compound activity in the NCI-60 Human Tumor Cell Line Screen. By comparing a compound’s activity fingerprint across the 60 cell lines to historical and contemporary data, COMPARE predicts potential mechanisms of action, tumor selectivity, and chemical analogs.

COMPARE has been an integral part of drug discovery at NCI for over 30 years and continues to evolve with the new HTS384 screen. 

What COMPARE Does

COMPARE identifies compounds with similar cell line activity patterns by calculating correlation coefficients between your compound and known reference agents.

You can:

  • Use a submitted NSC number as a “seed” compound
  • Compare it against:
    • Standard agents (e.g., paclitaxel, doxorubicin)
    • Synthetic compound database
    • Natural product extracts
    • Molecular Expression patterns (microarray and RnaSeq)
  • Rank results based on Pearson Correlation Coefficient (PCC)

The higher the PCC, the more likely two compounds share a common mechanism of action.  

Accessing COMPARE

Access TypeUse CaseLink
Public COMPAREAnalyze or review public patterns  Launch Public COMPARE 
Private COMPAREView private or embargoed data (DTP staff & suppliers only)Launch Private COMPARE 

How It Works

  1. Select a seed compound (usually your NSC)
  2. Choose dataset and correlation type
    • GI50, TGI, LC50 values
    • Pearson Correlation (standard method)
  3. COMPARE searches the database
    • Public: ~50,000+ compounds
    • Private: internal submissions, prepublication data
  4. Results display:
    • Ranked compound list
    • Similarity scores
    • Optional delta plots (mean graph-style visualizations) 

COMPARE Output Examples

  • Top Matches: Closest analogs in function
  • PCC Values: Degree of pattern similarity
  • Mean Graph Patterns: Cell line sensitivity profiles
  • Insights: Possible biochemical targets or resistance profiles 

COMPARE can be used to:

  • Hypothesize new mechanisms of action
  • Identify drug analogs with shared selectivity
  • Refine lead optimization in drug development
  • Connect screen results to molecular target data 

 

COMPARE and the HTS384 Screen

COMPARE remains compatible with both:

  • Legacy SRB-based screen data (1990–2023)
  • Current HTS384 luminescence-based data

When submitting new compounds through the HTS384 platform, your results can be analyzed via COMPARE once screening is complete.

Need help interpreting results? 

email moltarget@mail.nih.gov for assistance.

 Origin of COMPARE

COMPARE was developed in the early 1990s to harness the power of large-scale in vitro screening. Its signature innovation was the ability to correlate cell line sensitivity patterns, revealing potential shared mechanisms even among structurally unrelated compounds.

Landmark Contributions

  • Halichondrin B (marine compound) identified as a novel tubulin binder
  • Discovery of previously unknown topoisomerase II inhibitors
  • Reverse searches to connect mechanism to chemical novelty 

Legacy Methods (Archived)

Earlier versions of COMPARE supported two analysis modes:

  • Average Difference Method (ADM) – retired in 2025
  • Pearson Correlation Method (PCC) – currently used

COMPARE continues to evolve with improved normalization, subpanel scoring, and enhanced user interfaces.

 Integration with Molecular Target Data

COMPARE can be paired with molecular characterization data from the NCI-60 panel, allowing correlation between compound activity and target expression levels.

Visit the Molecular Targets Program to access or generate target data. 

Data

File DescriptionFile NameDownload Link
QC reports on sample MaterialAnalytical_Data_100_CompoundsAnalytical Data
NCI-60 concentration/response data and endpoint values for the IOA setoncologydrugsCompareOncologydrugscompare

 

If you would like to reproduce some or all of this content, see Reuse of NCI Information for guidance about copyright and permissions. In the case of permitted digital reproduction, please credit the National Cancer Institute as the source and link to the original NCI product using the original product's title; e.g., “History of the NCI-60 screen and COMPARE algorithm was originally published by the National Cancer Institute.”

Email