Generative and interpretable machine learning for aptamer design and analysis of in vitro sequence selection
importance 4/10
ai assisted
Advanced biosensing and bioanalysis techniques aptamer biotechnology stage: research
Problem
Obtaining aptamer binders via SELEX requires multiple rounds of experimental selection, and predicting which sequences will bind is difficult.
Why it matters
Aptamers are key diagnostic and therapeutic binders, and multi-round SELEX experiments are slow and costly.
What the AI did
Restricted Boltzmann Machines trained on single-round SELEX sequence ensembles score sequences by fitness, predict effects of selection at later rounds, and generate candidate binders.
Results
not stated in source
Impact
The method could reduce experimental SELEX rounds by predicting selection outcomes from early-round data.
Record
| Venue | PLoS Computational Biology |
|---|---|
| Published | 2022 |
| Work type | article |
| Authors | Andrea Di Gioacchino, Jonah Procyk, Marco Molari, John S. Schreck, Yu Zhou, Yan Liu, Rémi Monasson, Simona Cocco et al. |
| Citations | 41 (OpenAlex, at import time) |
| References | 64 |
| Model or method | Restricted Boltzmann Machines (RBMs) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1371/journal.pcbi.1010561 |
| Open access | gold |
| Keywords | aptamer, SELEX, RBM, generative model, thrombin |
Importance rating
4
real but narrower advance
Aptamers are diagnostic/therapeutic binders and multi-round SELEX is slow and costly, but the abstract frames a narrow methodological advance without scale evidence.