DeLA-Drug: A Deep Learning Algorithm for Automated Design of Druglike Analogues

Creanza, Teresa Maria; Lamanna, Giuseppe; Delre, Pietro; Contino, Marialessandra; Corriero, Nicola; Mangiatordi, Giuseppe Felice; Ancona, Nicola

doi:10.1021/acs.jcim.2c00205

Cited by 23 publications

(24 citation statements)

References 63 publications

(107 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to the target-directed RNN, the molecules resulting from SMILES-based exploration were on average more similar to the original compounds. This similarity is comparable to the similarity of the SMILESbased analog generator created by Creanza et al [60]. The distribution of scaffolds generated by SMILES-based exploration was also more similar to the known ligands.…”

Section: Exploration Of Nearby Chemical Spacesupporting

confidence: 77%

“…This was due to the fact that in most cases the original molecule was regenerated. In a recent study, Creanza et al created a SMILES-based analog generator that works by sampling SMILES with substitutions [60]. When using 5 substitutions, this approach has a lower validity (11 %) but higher uniqueness (60 %) compared to SMILES correction for exploration.…”

Section: Exploration Of Nearby Chemical Spacementioning

confidence: 99%

See 1 more Smart Citation

UnCorrupt SMILES: a novel approach to de novo design

Schoenmaker

Béquignon

Jespers

et al. 2022

Preprint

View full text Add to dashboard Cite

Generative deep learning models have emerged as a powerful approach for de novo drug design, as they aid researchers in finding new molecules with desired properties. Despite continuous improvements in the field, a subset of the outputs that sequence-based de novo generators produce cannot be progressed due to errors. Here, we propose to fix these invalid outputs post hoc. In similar tasks, transformer models from the field of natural language processing have been shown to be very effective. Therefore, here this type of model was trained to translate invalid Simplified Molecular-Input Line-Entry System (SMILES) into valid representations. The performance of this SMILES corrector was evaluated on four representative methods of de novo generation: a recurrent neural network (RNN), a target-directed RNN, a generative adversarial network (GAN), and a variational autoencoder (VAE). This study has found that the percentage of invalid outputs from these specific generative models ranges between 4 and 89 %, with different models having different error type distributions. Post hoc correction of SMILES increases model validity, with the SMILES corrector fixing 35 to 80 % of invalid model outputs. While, corrector models trained with one error per input sequence alter 60 to 90 % of invalid inputs, a higher performance was obtained for transformer models trained with multiple errors per input. In this case, the best model was able to correct 60 to 95 % of invalid generator outputs. Further analysis showed that these fixed molecules are comparable to the correct molecules from the de novo generators with regard to novelty and similarity. Additionally, the SMILES corrector can also be used to expand the amount of interesting new molecules within the targeted chemical space. Introducing different errors into existing molecules yields novel analogs with a uniqueness of 39 % and a novelty of approximately 20 %. The results of this research demonstrate that SMILES correction is a viable post hoc extension and can enhance the search for better drug candidates.

show abstract

Section: Exploration Of Nearby Chemical Spacesupporting

confidence: 77%

Section: Exploration Of Nearby Chemical Spacementioning

confidence: 99%

UnCorrupt SMILES: a novel approach to de novo design

Schoenmaker

Béquignon

Jespers

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The virtual drug screening can also be used for de novo compounds by integrating the deep learning-based de novo compounds generative model ( Gupta et al, 2018 ). Many more advanced deep learning generative models based on RNN architecture have been gradually developed ( Moret et al, 2021 ; Creanza et al, 2022 ). Among them, the beam search algorithm proposed by Michael Moret, et al, has great potential to generate more reasonable novel compounds ( Moret et al, 2021 ).…”

Section: Introductionmentioning

confidence: 99%

“…And they have tried to better adapt them to VS procedures and take into account some of VS limitations. Also, a new strategy called sampling with substitutions (SWS) usually generates molecules structurally similar to bioactive compounds or with given desired properties ( Creanza et al, 2022 ). It generates molecules structurally similar to bioactive compounds or with given desired properties.…”

Section: Introductionmentioning

confidence: 99%

An Efficient Modern Strategy to Screen Drug Candidates Targeting RdRp of SARS-CoV-2 With Potentially High Selectivity and Specificity

et al. 2022

View full text Add to dashboard Cite

Desired drug candidates should have both a high potential binding chance and high specificity. Recently, many drug screening strategies have been developed to screen compounds with high possible binding chances or high binding affinity. However, there is still no good solution to detect whether those selected compounds possess high specificity. Here, we developed a reverse DFCNN (Dense Fully Connected Neural Network) and a reverse docking protocol to check a given compound’s ability to bind diversified targets and estimate its specificity with homemade formulas. We used the RNA-dependent RNA polymerase (RdRp) target as a proof-of-concept example to identify drug candidates with high selectivity and high specificity. We first used a previously developed hybrid screening method to find drug candidates from an 8888-size compound database. The hybrid screening method takes advantage of the deep learning-based method, traditional molecular docking, molecular dynamics simulation, and binding free energy calculated by metadynamics, which should be powerful in selecting high binding affinity candidates. Also, we integrated the reverse DFCNN and reversed docking against a diversified 102 proteins to the pipeline for assessing the specificity of those selected candidates, and finally got compounds that have both predicted selectivity and specificity. Among the eight selected candidates, Platycodin D and Tubeimoside III were confirmed to effectively inhibit SARS-CoV-2 replication in vitro with EC50 values of 619.5 and 265.5 nM, respectively. Our study discovered that Tubeimoside III could inhibit SARS-CoV-2 replication potently for the first time. Furthermore, the underlying mechanisms of Platycodin D and Tubeimoside III inhibiting SARS-CoV-2 are highly possible by blocking the RdRp cavity according to our screening procedure. In addition, the careful analysis predicted common critical residues involved in the binding with active inhibitors Platycodin D and Tubeimoside III, Azithromycin, and Pralatrexate, which hopefully promote the development of non-covalent binding inhibitors against RdRp.

show abstract

“…To our knowledge, there still no LSTM based model for peptide generation and finetuning to obtain de novo active peptide. However, there are much more existing small-molecule generative models for small molecules [18][19][20], and small molecules with specific potential biological activities can be targeted by finetuning. Similar to the generation of de novo compounds, with the increasing accumulation of active biological peptides as a training set, it is now possible to use deep learning models to generate many potentially biologically active peptides.…”

Section: Introductionmentioning

confidence: 99%

Deep-learning based bioactive therapeutic peptides generation and screening

Zhang¹,

Saravanan

Wei

et al. 2022

Preprint

View full text Add to dashboard Cite

Many bioactive peptides demonstrated therapeutic effects over-complicated diseases, such as antiviral, antibacterial, anticancer, etc. Similar to the generating de novo chemical compounds, with the accumulated bioactive peptides as a training set, it is possible to generate abundant potential bioactive peptides with deep learning. Such techniques would be significant for drug development since peptides are much easier and cheaper to synthesize than compounds. However, there are very few deep learning-based peptide generating models. Here, we have created an LSTM model (named LSTM_Pep) to generate de novo peptides and finetune learning to generate de novo peptides with certain potential therapeutic effects. Remarkably, the Antimicrobial Peptide Database has fully utilized in this work to generate various kinds of potential active de novo peptide. We proposed a pipeline for screening those generated peptides for a given target, and use Main protease of SARS-COV-2 as concept-of-proof example. Moreover, we have developed a deep learning-based protein-peptide prediction model (named DeepPep) for fast screening the generated peptides for the given targets. Together with the generating model, we have demonstrated iteratively finetune training, generating and screening peptides for higher predicted binding affinity peptides can be achieved. Our work sheds light on to the development of deep learning-based methods and pipelines to effectively generating and getting bioactive peptides with a specific therapeutic effect, and showcases how artificial intelligence can help discover de novo bioactive peptides that can bind to a particular target.

show abstract

DeLA-Drug: A Deep Learning Algorithm for Automated Design of Druglike Analogues

Cited by 23 publications

References 63 publications

UnCorrupt SMILES: a novel approach to de novo design

UnCorrupt SMILES: a novel approach to de novo design

An Efficient Modern Strategy to Screen Drug Candidates Targeting RdRp of SARS-CoV-2 With Potentially High Selectivity and Specificity

Deep-learning based bioactive therapeutic peptides generation and screening

Contact Info

Product

Resources

About