Related Experiment Video
Updated: Oct 11, 2025
12:37
Efficient Nucleic Acid Extraction and 16S rRNA Gene Sequencing for Bacterial Community Characterization
Published on: April 14, 2016
39.2K
Understanding Data Noise and Uncertainty through Analysis of Replicate Samples in DNA-Encoded Library Selection
Hongyao Zhu1, Timothy L Foley2, Justin I Montgomery2
1Simulation and Modelling Sciences, Pfizer Inc., Groton, Connecticut 06340, United States.
Journal of Chemical Information and Modeling
|December 6, 2021
Summary
Researchers developed a method to estimate experimental noise in DNA-Encoded Library (DEL) selections. This approach improves data quality assessment and analysis by normalizing replicates and identifying noise factors like sequencing depth.
Area of Science:
- Drug Discovery
- Biotechnology
- Computational Biology
Background:
- DNA-Encoded Library (DEL) technology enables rapid screening of vast chemical libraries.
- Accurate assessment of experimental noise is crucial for reliable DEL data interpretation.
- Existing methods may struggle with high-throughput data and variability in selection conditions.
Purpose of the Study:
- To develop a robust method for estimating noise levels in DEL experiments.
- To provide a data-driven approach for assessing DEL data quality.
- To understand factors influencing noise in DEL selections.
Main Methods:
- Analysis of replicate DNA-Encoded Library (DEL) selection datasets.
- Application of logarithm transformation to compound counts.
- Normalization of replicate data to estimate noise.
- Investigation of noise dependency on sequencing depth and selection conditions.
Main Results:
- A novel approach for noise level estimation in DEL experiments was successfully developed.
- Noise estimation is independent of compound frequency cutoffs.
- Removing low-frequency compounds (1-5 read counts) significantly reduces dataset size without affecting results.
- Noise levels are influenced by sequencing depth and specific experimental conditions.
Conclusions:
- The developed method provides a reliable way to assess DEL experimental quality.
- Noise estimation aids in more accurate hit identification and data interpretation.
- Data preprocessing strategies, like removing low-count compounds, can enhance DEL analysis efficiency.
