Propedia v2.3: A novel representation approach for the peptide-protein interaction database using graph-based structural signatures

Martins, Pedro; Mariano, Diego; Carvalho, Frederico Chaves; Bastos, Luana Luiza; Moraes, Lucas Brunelli de; Paixão, Vivian; Melo‐Minardi, Raquel Cardoso de

doi:10.3389/fbinf.2023.1103103

Cited by 9 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The number of final screened complex instances of peptide-protein pairs is 9070. Further, we screen the existing datasets of PropediaV2.3 (Martins et al, 2023) and PepBDB (Wen et al, 2018) with the same criterion, leading to additional 6523 instances to expand it. Appendix.…”

Section: Datasetmentioning

confidence: 99%

“…For training the deep learning model, we collect a benchmark dataset called PPBench2024 consisting of 15593 protein-peptide pairs with high quality. Compared with previous protein-peptide datasets (Agrawal et al, 2019;Martins et al, 2023;Wen et al, 2018), PPBench2024 augments and expands available data, and filters large amounts of low-quality data according to strict criteria, thus allowing deep-learning-based models to be trained on the datasets and to fulfill the peptide drug discovery tasks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

PPFlow: Target-Aware Peptide Design with Torsional Flow Matching

Lin,

Zhang,

Zhao

et al. 2024

Preprint

View full text Add to dashboard Cite

Therapeutic peptides have proven to have great pharmaceutical value and potential in recent decades. However, methods of AI-assisted peptide drug discovery are not fully explored. To fill the gap, we propose a target-aware peptide design method called PPFlow, based on conditional flow matching on torus manifolds, to model the internal geometries of torsion angles for the peptide structure design. Besides, we establish a protein-peptide binding dataset namedPPBench2024to fill the void of massive data for the task of structure-based peptide drug design and to allow the training of deep learning methods. Extensive experiments show that PPFlowreaches state-of-the-art performance in tasks of peptide drug generation and optimization in comparison with baseline models, and can be generalized to other tasks including docking and side-chain packing.

show abstract

Section: Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

PPFlow: Target-Aware Peptide Design with Torsional Flow Matching

Lin,

Zhang,

Zhao

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Owing to their increased flexibility and low toxicity, peptides have become a focal point in medical research, particularly for pharmaceutical applications aimed at treating various diseases, including chronic inflammatory conditions. Notable peptide databases include PeptiDB [89], PepX [90], and Propedia [91,92].…”

Section: D-structuresmentioning

confidence: 99%

The Role of Structural Bioinformatics in Understanding Tumor Necrosis Factor α-Interacting Protein Mechanisms in Chronic Inflammatory Diseases: A Review

Bastos,

Mariano,

Lemos

et al. 2024

Immuno

Self Cite

View full text Add to dashboard Cite

Tumor necrosis factor α (TNF-α) is a multifunctional cytokine protein acknowledged as a vital mediator in cell differentiation, proliferation, and survival. Additionally, TNF-α is a crucial component of the host’s defense by mediating inflammatory and immune responses against various aggressive agents, including viruses, bacteria parasites, and tumors. However, excessive production can be detrimental to the body and is also implicated in developing several inflammatory and immune-mediated disorders. Therefore, there is great interest in studying its role and its modulation, in various diseases, both in in vitro, in vivo, and in silico experiments. In this review, we evaluated the structures of proteins related to TNF-α available in public databases. In addition, we described the main antibodies blocking this cytokine and its applications and commented on the potential of naturally produced binding molecules, such as TNF-α-binding proteins produced by ticks. We also discuss the role of structural bioinformatics techniques in understanding the mechanisms of chronic inflammatory diseases related to TNF-α. We hope that the data presented in this review will be useful for studies that aim to better understand the mechanisms of the interactions of TNF-α with other proteins and will lead to new drugs or treatments.

show abstract

“…In this context, peptides have emerged as an alternative for inhibiting the causative agent. Peptides are molecules composed of short amino acid residues of 2-50 sequences linked by peptide bonds [15,16]. They play crucial roles in living beings, including cell signaling and immune modulation [17,18].…”

Section: Introductionmentioning

confidence: 99%

An Approach for Engineering Peptides for Competitive Inhibition of the SARS-COV-2 Spike Protein

de Abreu,

Carvalho,

Mariano

et al. 2024

Molecules

Self Cite

View full text Add to dashboard Cite

SARS-CoV-2 is the virus responsible for a respiratory disease called COVID-19 that devastated global public health. Since 2020, there has been an intense effort by the scientific community to develop safe and effective prophylactic and therapeutic agents against this disease. In this context, peptides have emerged as an alternative for inhibiting the causative agent. However, designing peptides that bind efficiently is still an open challenge. Here, we show an algorithm for peptide engineering. Our strategy consists of starting with a peptide whose structure is similar to the interaction region of the human ACE2 protein with the SPIKE protein, which is important for SARS-COV-2 infection. Our methodology is based on a genetic algorithm performing systematic steps of random mutation, protein–peptide docking (using the PyRosetta library) and selecting the best-optimized peptides based on the contacts made at the peptide–protein interface. We performed three case studies to evaluate the tool parameters and compared our results with proposals presented in the literature. Additionally, we performed molecular dynamics (MD) simulations (three systems, 200 ns each) to probe whether our suggested peptides could interact with the spike protein. Our results suggest that our methodology could be a good strategy for designing peptides.

show abstract

Propedia v2.3: A novel representation approach for the peptide-protein interaction database using graph-based structural signatures

Cited by 9 publications

References 16 publications

PPFlow: Target-Aware Peptide Design with Torsional Flow Matching

PPFlow: Target-Aware Peptide Design with Torsional Flow Matching

The Role of Structural Bioinformatics in Understanding Tumor Necrosis Factor α-Interacting Protein Mechanisms in Chronic Inflammatory Diseases: A Review

An Approach for Engineering Peptides for Competitive Inhibition of the SARS-COV-2 Spike Protein

Contact Info

Product

Resources

About