AptaNet as a deep learning approach for aptamer–protein interaction prediction
Problem
Identifying which short DNA/RNA aptamers bind to which target proteins is slow and costly to do experimentally. AptaNet was built to predict aptamer-protein interaction pairs computationally.
Why it matters
Aptamers are key to biosensing, therapeutics, and diagnostics, so faster pair identification could cut cost and time.
What the AI did
AptaNet, a deep neural network, integrates k-mer aptamer encodings and protein composition features, handles class imbalance with a neighborhood cleaning algorithm, and uses random forest feature selection to predict aptamer-protein pairs.
Results
not stated in source (abstract truncated mid-result at '99.')
Impact
Could speed aptamer selection for biosensing and therapeutic applications, per the authors' claims.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Neda Emami, Reza Ferdousi |
| Citations | 65 (OpenAlex, at import time) |
| References | 121 |
| Model or method | AptaNet (deep neural network) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41598-021-85629-0 |
| Open access | gold |
| Keywords | aptamer, protein interaction, deep learning, prediction, neural network |
Importance rating
real but narrower advance
Aptamers have stated potential in biosensing, therapeutics, diagnostics; the tool predicts interaction pairs on a benchmark dataset, a narrower computational advance with no stated scale.