Sequence-based prediction of drug–target binding using machine learning, deep learning and ensemble models without 3D structural information
Nazife Çevik, Taner Çevik, Ahmet Gürhanlı · 2026
Read Sequence-based prediction of drug–target binding using machine learning, deep learning and ensemble models without 3D structural information by Nazife Çevik, Taner Çevik, Ahmet Gürhanlı (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13518898.txt.
Description
Nazife Çevik, Taner Çevik, and Ahmet Gürhanlı show in 2026 that drug-target binding can be predicted from sequences alone, without any 3D protein structure. Early drug discovery needs to know which molecules bind which proteins, while docking methods depend on structures that are often missing or shaky.
Their framework merges chemical descriptors of proteins, short sequence motifs, and sequence-like drug codes into one feature space. Many model types were trained with imbalance correction, and a stacked ensemble combining three classifiers performed best, passing 0.90 in ROC-AUC with a peak of 0.914. Feature analysis surfaced biologically sensible binding motifs. Docking checks on a subset agreed with the predictions, supporting the sequence-only route.
Catalog record
| CREATOR | Nazife Çevik, Taner Çevik, Ahmet Gürhanlı |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63601 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |