TY - JOUR
T1 - Generative and interpretable machine learning for aptamer design and analysis of in vitro sequence selection
AU - Di Gioacchino, Andrea
AU - Procyk, Jonah
AU - Molari, Marco
AU - Schreck, John S.
AU - Zhou, Yu
AU - Liu, Yan
AU - Monasson, Rémi
AU - Cocco, Simona
AU - Šulc, Petr
N1 - Funding Information:
This work was supported by National Science Foundation (No. 2155095 and TGBIO210009 to P.S.), Agence Nationale de la Recherche (RBMPro CE30-0021-01 and ANR-19 Decrypted CE30-0021-01 to S.C. and R.M.), and European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement (No 101026293 to A.D.G.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Publisher Copyright:
Copyright: © 2022 Di Gioacchino et al.
PY - 2022/9
Y1 - 2022/9
N2 - Selection protocols such as SELEX, where molecules are selected over multiple rounds for their ability to bind to a target of interest, are popular methods for obtaining binders for diagnostic and therapeutic purposes. We show that Restricted Boltzmann Machines (RBMs), an unsupervised two-layer neural network architecture, can successfully be trained on sequence ensembles from single rounds of SELEX experiments for thrombin aptamers. RBMs assign scores to sequences that can be directly related to their fitnesses estimated through experimental enrichment ratios. Hence, RBMs trained from sequence data at a given round can be used to predict the effects of selection at later rounds. Moreover, the parameters of the trained RBMs are interpretable and identify functional features contributing most to sequence fitness. To exploit the generative capabilities of RBMs, we introduce two different training protocols: one taking into account sequence counts, capable of identifying the few best binders, and another based on unique sequences only, generating more diverse binders. We then use RBMs model to generate novel aptamers with putative disruptive mutations or good binding properties, and validate the generated sequences with gel shift assay experiments. Finally, we compare the RBM’s performance with different supervised learning approaches that include random forests and several deep neural network architectures.
AB - Selection protocols such as SELEX, where molecules are selected over multiple rounds for their ability to bind to a target of interest, are popular methods for obtaining binders for diagnostic and therapeutic purposes. We show that Restricted Boltzmann Machines (RBMs), an unsupervised two-layer neural network architecture, can successfully be trained on sequence ensembles from single rounds of SELEX experiments for thrombin aptamers. RBMs assign scores to sequences that can be directly related to their fitnesses estimated through experimental enrichment ratios. Hence, RBMs trained from sequence data at a given round can be used to predict the effects of selection at later rounds. Moreover, the parameters of the trained RBMs are interpretable and identify functional features contributing most to sequence fitness. To exploit the generative capabilities of RBMs, we introduce two different training protocols: one taking into account sequence counts, capable of identifying the few best binders, and another based on unique sequences only, generating more diverse binders. We then use RBMs model to generate novel aptamers with putative disruptive mutations or good binding properties, and validate the generated sequences with gel shift assay experiments. Finally, we compare the RBM’s performance with different supervised learning approaches that include random forests and several deep neural network architectures.
UR - http://www.scopus.com/inward/record.url?scp=85139739704&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85139739704&partnerID=8YFLogxK
U2 - 10.1371/journal.pcbi.1010561
DO - 10.1371/journal.pcbi.1010561
M3 - Article
C2 - 36174101
AN - SCOPUS:85139739704
SN - 1553-734X
VL - 18
JO - PLoS computational biology
JF - PLoS computational biology
IS - 9
M1 - e1010561
ER -