Practical tools and resources for turning drug discovery data into useful information.

Data analysis is an essential part of the drug discovery process. Modern screening and biophysical experiments can generate large and complex datasets, and the way those data are organized, normalized, visualized, and interpreted can have a major impact on the conclusions that can be drawn.

The DDRC develops practical computational tools and workflows to help researchers analyze their experimental data. We also highlight publicly available software that we find useful for drug discovery research.

This page brings these resources together in one place. We will continue to add tools, examples, and tips and tricks based on our experience working with drug discovery data.

DDRC Tools

HTS Normalization Tool

From raw screening data to analysis-ready results.

The DDRC HTS Normalization Tool is a Python-based tool for processing high-throughput screening data. It combines raw screening measurements with a compound database, links compounds to their corresponding plate and well positions, performs normalization, and generates quality-control and screening results.

The tool calculates:

• Percent inhibition
• Raw data z-score — plate-based
• Percent inhibition z-score — plate-based
• Run percent inhibition z-score — run-based
• Z′ factor for each plate
• Plate-level quality-control information
• Potential hits based on a user-defined percent-inhibition threshold
• Compound identifiers and structures linked to screening results

The program accepts a screening dataset and a separate compound database and uses Plate + Well as the link between the two. Compound structures are retained as SMILES/CXSMILES, allowing downstream chemical analysis and visualization.

The tool is designed to be straightforward to run from the command line and to provide a transparent, reproducible workflow that researchers can inspect and modify for their own applications.

The tool is publicly posted on GitHub. A formal open-source license and release are being deferred pending review by
the Rockefeller University Office of Technology Transfer.

Download HTS Normalization Tool v1.0

High-Throughput Concentration-Response Tool (soon to be released)

Purpose and Overview
The DDRC DRC Analysis Tool is a Python-based utility for automated concentration-response analysis. It processes dose-response data, fits four-parameter logistic (4PL) curves, evaluates curve quality, and produces an Excel workbook containing compound-level results and graphical representations of the fitted curves.
Principal Capabilities
• Accepts CSV, XLSX, XLS, and ODS input data.
• Provides interactive mapping of plate, well, compound, concentration, response, and optional batch fields.
• Supports column-, row-, and specific-well control arrangements, as well as datasets in which responses are already
expressed as percent inhibition.
• Automatically handles replicate measurements and can overlay a reference-compound concentration-response
curve.
• Fits 4PL concentration-response curves and reports IC50, chi-square, AUC, curve class, fit confidence, DDRC Curve
Score, flags, and fit status.
• Generates compound-level plots and plate heatmaps and embeds plots directly into the Results worksheet of the Excel output.

Inputs
Experimental concentration-response data supplied as CSV, Excel, or OpenDocument spreadsheets. The program
interactively identifies the relevant columns and allows users to specify the control arrangement and, optionally, a
reference compound.

Outputs
An Excel workbook containing compound-level curve-analysis results, fitted-curve plots, plate heatmaps, fit-status information, and quality metrics including the DDRC Curve Score.
Software EnvironmentPython. Direct dependencies specified in requirements.txt: pandas, NumPy, matplotlib, SciPy, openpyxl, odfpy, and xlrd. These packages are not bundled with the software distribution and are installed separately by the end user.

Cheminformatics and Medicinal Chemistry Tools (soon to be released).

Recommended Tools

DataWarrior

Chemical and biological data visualization and analysis.

DataWarrior is a freely available platform for exploring chemical and biological datasets.

DataWarrior is particularly useful for analyzing screening results after normalization. Screening data containing compound structures as SMILES or CXSMILES can be imported and explored alongside experimental measurements. Chemical properties, structural similarities, activity relationships, and other molecular characteristics can then be examined interactively.

One advantage of this approach is that the HTS normalization workflow does not need to calculate and store every possible molecular descriptor. The structure itself can be retained in the screening dataset, while downstream software can calculate the properties needed for a particular analysis.

[Learn more about DataWarrior]

Tips & Tricks

Drug discovery data can be surprisingly difficult to organize and interpret. Small decisions about data structure, normalization, quality control, and visualization can have a large impact on the usefulness of the final results.

We are developing a collection of practical tips, workflows, and lessons learned from working with real experimental datasets.

Understanding z-scores in HTS

What is the difference between a plate-based and run-based z-score? When should each be used, and what does a z-score actually tell you about an individual well?

Coming soon

Organizing Screening Data

A simple approach to organizing screening data using Plate + Well + Compound ID + Structure + Measurement can make downstream analysis much easier.

Coming soon

Choosing the Right Data Format

CSV, Excel, and modern columnar formats such as Parquet each have advantages. Understanding when to use each can make working with large screening datasets considerably easier.

Coming soon

Need Help With Your Data?

Not sure how to analyze your screening or assay data?

The DDRC can help researchers develop practical approaches for organizing, analyzing, visualizing, and interpreting data generated from drug discovery experiments.

[Contact the DDRC]