Abstract:DNA-encoded library (DEL) is a powerful ligand discovery technology that has been widely adopted in the pharmaceutical industry. DEL selections are typically performed with a purified protein target immobilized on a matrix or in solution phase. Recently, DELs have also been used to interrogate the targets in the complex biological environment, such as membrane proteins on live cells. However, due to the complex landscape of the cell surface, the selection inevitably involves significant nonspecific interactions, and the selection data are much noisier than the ones with purified proteins, making reliable hit identification highly challenging. Researchers have developed several approaches to denoise DEL datasets, but it remains unclear whether they are suitable for cell-based DEL selections. Here, we report the proof-of-principle of a new machine-learning (ML)-based approach to process cell-based DEL selection datasets by using a Maximum A Posteriori (MAP) estimation loss function, a probabilistic framework that can account for and quantify uncertainties of noisy data. We applied the approach to a DEL selection dataset, where a library of 7,721,415 compounds was selected against a purified carbonic anhydrase 2 (CA-2) and a cell line expressing the membrane protein carbonic anhydrase 12 (CA-12). The extended-connectivity fingerprint (ECFP)-based regression model using the MAP loss function was able to identify true binders and also reliable structure-activity relationship (SAR) from the noisy cell-based selection datasets. In addition, the regularized enrichment metric (known as MAP enrichment) could also be calculated directly without involving the specific machine-learning model, effectively suppressing low-confidence outliers and enhancing the signal-to-noise ratio. Future applications of this method will focus on de novo ligand discovery from cell-based DEL selections.

Agonist Discovery for Membrane Proteins on Live Cells by Using DNA-encoded Libraries

Selection of DNA-encoded chemical libraries against endogenous membrane proteins on live cells

Direct Ligand Screening Against Membrane Proteins On Live Cells Enabled By Dna-Programmed Affinity Labelling

Protein-templated Ligand Discovery Via the Selection of DNA-encoded Dynamic Libraries

Selecting a DNA-Encoded Chemical Library Against Non-immobilized Proteins Using a “Ligate–cross-Link–purify” Strategy

A Proximity Based General Method for Identification of Ligand and Receptor Interactions in Living Cells.

Small-molecule discovery through DNA-encoded libraries

Selecting Agonists from Single Cells Infected with Combinatorial Antibody Libraries.

Discovering Cell-targeting Ligands and the Cell Surface Receptors via Selection of DNA-encoded Chemical Libraries (DELs) against Cancer Cells without Predefined Targets

Strategies for developing DNA-encoded libraries beyond binding assays

Machine-Learning-Based Data Analysis Method for Cell-Based Selection of DNA-Encoded Libraries

Selection of DNA‐Encoded Dynamic Chemical Libraries for Direct Inhibitor Discovery

Dna-Encoded Dynamic Chemical Library And Its Applications In Ligand Discovery

DNA-Encoded Library Technology─A Catalyst for Covalent Ligand Discovery

Cover Picture: Selection of DNA‐Encoded Small Molecule Libraries Against Unmodified and Non‐Immobilized Protein Targets (angew. Chem. Int. Ed. 38/2014)

Selection of DNA‐Encoded Small Molecule Libraries Against Unmodified and Non‐Immobilized Protein Targets

Force-Enhanced Sensitive Detection of New DNA-Interactive Agents from Microorganisms at the Single-Molecule Level

Guiding Protein Delivery into Live Cells Using DNA-programmed Membrane Fusion.

ABPP-CoDEL: Activity-Based Proteome Profiling-Guided Discovery of Tyrosine-Targeting Covalent Inhibitors from DNA-Encoded Libraries

A Membrane-Confined Signal Amplification Strategy for Sensitive Monitoring of Extracellular Enzymatic Activity Upon Drug Stimulus.

A Proximity-Based Assay for Identification of Ligand and Membrane Protein Interaction in Living Cells.